Missing American child FOUND ALIVE in Mexico 25 Years Later | Cold Case Solved

Missing American child FOUND ALIVE in Mexico 25 Years Later | Cold Case Solved

We have new developments tonight in a decades-old cold case out of New Haven. DNA evidence has confirmed that a woman living in Mexico was kidnapped from New Haven back in 1999. In October of 1999, a 23-month-old girl named Andrea Michelle Reyes was dropped off in New Haven, Connecticut for a scheduled visitation with her mother. It was routine. It was court-ordered. It was the kind of arrangement thousands of separated families navigate every single week. Andrea’s father, Carlos Reyes, had legal custody. Her mother, Rosa Tenorio, did not. Rosa never brought her back. By the time anyone realized what had happened, Rosa Tenorio and her daughter were already gone, out of Connecticut, out of the country, headed for Mexico.

A felony warrant for custodial interference was issued against Rosa Tenorio before 1999 was even over. The FBI got involved. The National Center for Missing and Exploited Children opened a file. Age progression images were generated every few years, with artists rendering what Andrea might look like as a 5-year-old, a 10-year-old, a teenager, a young woman—pure speculation because nobody had a single confirmed sighting to work from. Here is the detail that should stop you cold: in the year 2000, just months after Andrea disappeared, Mexican authorities located Rosa Tenorio and her daughter. They found them. And according to a New Haven police sergeant speaking 25 years later, officials were advised that the government of Mexico would not take further action to remove Andrea from her mother’s custody.

The trail did not go cold because nobody could find them. It went cold because someone found them and chose not to act. Andrea Reyes grew up in the city of Puebla, Mexico, a metropolis of over a million and a half people southeast of Mexico City, under her mother’s care for the next 25 years. Her father never stopped looking. He made multiple trips to Mexico himself, searching in person for a daughter he had no way of knowing was alive. In 2023, a newly appointed New Haven police detective named Keelin Neikov pulled Andrea’s file and decided the case was not as dead as everyone assumed. Using search warrants, old interview transcripts, and social media, she found a woman living in Puebla who matched. She reached out. That woman agreed to submit her DNA.

In March of 2025, a forensic laboratory confirmed what Detective Neikov had suspected. The woman in Puebla was Andrea Michelle Reyes, missing for 25 years, 4 months, and some days. She spoke to her father for the first time since she was 23 months old by phone, then by message. The National Center for Missing and Exploited Children is now helping the family plan the in-person reunion neither of them ever thought would actually happen. The warrant for Rosa Tenorio is still active. She is believed to still be living in Mexico. She has never been arrested because Mexico will not extradite a mother for taking her own biological child—the same legal reality investigators now believe let her walk free within months of her daughter’s abduction in the first place.

What happened in the 25 years between a routine custody visit in Connecticut and a DNA lab confirming a father and daughter? And why did Mexico know where Andrea was almost from the very beginning and never intervene? It starts with a family that was already fractured long before anyone ever went missing. Welcome back to Vanished Women and Children. If you are new here, subscribe and hit the notification bell right now. Drop a comment and tell me where in the world you are watching from. Now, let’s get into it. Warning, this is a true crime story. The following program contains real events that include violence and sensitive topics. Some viewers may find certain details disturbing. Viewer discretion is advised.

To understand how Andrea Reyes vanished for a quarter of a century and to understand exactly where the systems meant to protect her broke down, you first have to understand the custody arrangement she was born into and the particular vulnerabilities that exist anytime a family separation crosses an international border. New Haven, Connecticut sits on the Long Island Sound in the southern part of the state, home to Yale University, a historic manufacturing base, and a diverse, densely populated urban community that includes significant Puerto Rican and broader Latino populations, many with extended family ties across the Caribbean and Latin America. It is exactly the kind of American city where a custody arrangement between two parents with connections to different countries would not be unusual at all.

Families like the Reyes family were and remain a common thread in the fabric of a city built by generations of immigration and cross-border movement, which made the eventual outcome of this case—a mother disappearing across an international border with her daughter—a possibility that the family court system handling their custody arrangement may not have been fully equipped to anticipate or prevent. Carlos Reyes and Rosa Tenorio had a daughter together, Andrea Michelle Reyes, born in the winter of 1997 or 1998. By 1999, when Andrea was 23 months old, her parents’ relationship had already deteriorated to the point where the family court system had become involved. Carlos had been awarded legal custody of Andrea. Rosa had been granted visitation rights, a common arrangement that allows a non-custodial parent to maintain a relationship with their child while placing the primary responsibility for daily care with the custodial parent.

Family court custody proceedings of this kind are, in the vast majority of cases across the United States, resolved with exactly this structure: one parent granted primary legal and physical custody, the other granted scheduled time with the child. Both parents are expected to honor the arrangement in good faith because the alternative—a parent using visitation as an opportunity to disappear with a child—represents such a fundamental violation of trust that courts rarely build extensive safeguards against it into routine custody orders. There is no equivalent in most family court proceedings of the kind of security protocol that might exist at an international border crossing. The system runs on the assumption that a parent granted visitation will return the child because the overwhelming majority of parents do exactly that week after week, year after year without incident.

This is precisely the kind of arrangement that in the vast majority of cases functions exactly as intended. Millions of children across the United States move between two households every week under custody agreements just like this one, and the system generally works. But every custody arrangement carries within it a specific point of vulnerability: the moment of handoff, the window of time when a child is physically transferred from one parent’s care to another’s, resting entirely on the good faith of both parents to honor what a court has ordered. In October of 1999, that point of vulnerability is exactly where everything went wrong. Andrea was brought to a scheduled visitation with her mother, a routine handoff, the kind that had likely happened many times before without incident.

Rosa Tenorio had her daughter for what was supposed to be a limited court-sanctioned period of time. She did not return her. By the time it became clear that this was not simply a delay or a misunderstanding, Rosa Tenorio and 23-month-old Andrea were already gone. There was no confrontation, no dramatic chase, no crime scene in the way this channel often describes. There was simply a visitation that never ended, a mother who took advantage of court-ordered access to her own daughter and used it as the opening she needed to disappear entirely. New Haven police opened an investigation. A felony arrest warrant for custodial interference was issued against Rosa Tenorio before the year was out, the specific charge that applies when a parent, even a biological parent, violates a custody order by taking or retaining a child in defiance of the court’s ruling.

It is important to understand what this charge does and does not represent. Custodial interference acknowledges that a parent without legal custody has broken the law by taking a child, but it exists within a fundamentally different legal framework than kidnapping by a stranger. The consequences, the jurisdictional reach, and as this case would go on to demonstrate with brutal clarity, the willingness of foreign governments to cooperate with enforcement are all substantially different when the person who took the child is the child’s own mother. Investigators believed almost immediately that Rosa Tenorio had left the country, taking Andrea to Mexico. The FBI became involved given the interstate and now international nature of the case. The National Center for Missing and Exploited Children opened a file on Andrea, generating the kind of ongoing public awareness campaign that organization has become known for: flyers, database entries, and as the years passed, a series of computer-generated age progression images attempting to render what Andrea might look like as she grew from a toddler into a child and eventually into a teenager and young woman nobody had ever actually seen.

Carlos Reyes did not simply wait for law enforcement to solve this. According to police, he made multiple trips to Mexico himself over the following years, searching in person for a daughter who could have been anywhere across a country of over 100 million people, armed with nothing but hope and whatever leads investigators could share with him. This is the section of the Andrea Reyes case that deserves the most scrutiny and the detail that should reframe everything you think you know about why this case took 25 years to solve. According to New Haven Police Sergeant John Moore speaking at the March 2025 press conference confirming Andrea’s identification, Mexican authorities identified Rosa Tenorio and Andrea in the year 2000. That is not 25 years after the abduction. That is not even 5 years after. That is within roughly 12 months of Andrea being taken from Connecticut.

Sergeant Moore’s exact words delivered publicly at that press conference were that officials were advised the government of Mexico would not take further action to remove Andrea from her mother’s custody. Sit with that for a moment. Within a year of a 23-month-old girl being taken across an international border in direct violation of a United States custody order and an active felony warrant, the country she had been taken to had already located her. And a decision made somewhere within the machinery of the Mexican government, for reasons that have never been publicly detailed in the reporting on this case, was that no further action would be taken to reunite her with the father who had legal custody and who was at that very moment searching desperately for her. This is not a story about a mother who was clever enough to permanently evade detection. Rosa Tenorio was found. The system worked in the narrowest possible sense exactly once within the first year, and then it simply stopped.

The reasons behind Mexico’s decision not to intervene in 2000 are not fully documented in public reporting, but the pattern this case reveals—one that would repeat itself 25 years later when Rosa Tenorio’s active arrest warrant proved impossible to enforce because Mexico will not extradite a biological mother—suggests a consistent legal and diplomatic reality. International custody disputes involving a child’s own biological parent occupy a strange and often under-enforced legal space, one where a country’s reluctance to remove a child from a biological parent, however legally unauthorized that parent’s custody may be under a foreign court order, can override the enforcement mechanisms that exist on paper. Carlos Reyes did not know in the year 2000 that his daughter had already been found and left in place. He continued searching for 23 more years, making his own trips to Mexico, working with the FBI and NCMEC, watching age progression images render a daughter he had no way of actually locating, all while somewhere in a government file, the answer to where Andrea was already existed.

Andrea Reyes grew up in Puebla, Mexico, a large historic city of more than one and a half million people located southeast of Mexico City, known for its colonial architecture, its culinary traditions, and its status as one of the country’s major urban centers. She grew up there under her mother’s care for the entirety of her childhood and into her adult life. Meanwhile, in Connecticut, the case followed the trajectory that far too many international parental abduction cases follow once the immediate search window closes: continued but increasingly quiet institutional effort, punctuated by periodic renewals of legal paperwork that served as evidence the case was technically still open, even as any realistic hope of a breakthrough faded with each passing year. A second felony warrant was issued for Rosa Tenorio in 2009, a full decade after the original abduction, and a clear indication that the case, while cold in practice, had never been formally closed by New Haven authorities. Andrea’s information remained entered in the National Missing and Unidentified Persons System as case number MP3324, sitting alongside thousands of other unresolved missing persons cases across the country.

The age progression images continued to be generated periodically, a technique that attempts to use software and forensic art techniques to estimate how a missing child’s face might change as they age based on the facial structures of their known biological relatives. These images represented for over two decades essentially the only tool investigators had to work with in a case with no confirmed sightings, no tips that led anywhere concrete, and a suspect who had apparently vanished into a country of over 120 million people. Consider what it means to look at a computer-generated approximation of your own daughter’s face rendered by software attempting to guess how a toddler’s features might mature across a decade and then another decade, knowing that the actual woman those images are attempting to represent could be walking down any street in any city in an entire country, utterly unaware that anyone was looking for her at all—or worse, aware on some level that something about her own history did not fully add up, but with no framework for understanding what that feeling meant.

This is the specific protracted form of grief that defines international parental abduction cases in a way that domestic cases, however devastating, rarely replicate. There is no local community to organize search parties. There is no reasonable expectation that a billboard or a flyer distributed in the right neighborhood might produce a sighting. The scale of the search is a nation away, filtered through language barriers, differing legal systems, and diplomatic channels that move when they move at all with agonizing slowness. For Carlos Reyes, these were two decades spent in the specific grinding uncertainty that defines so many of the cases this channel covers: not knowing whether his daughter was alive or dead, whether she was safe or in danger, whether she even remembered that she had ever had another name, another father, another life before Puebla. He built whatever life he could build around that absence. He continued, by every account, to hold on to hope. He remarried and built a family that included a stepmother who would 25 years later stand beside him at a press conference and speak about the daughter she had never met as though she had always been part of the family, because in the specific sustained way that a family holds space for someone missing, she always had been.

Cold cases do not solve themselves. They require someone—often, as in so many of the cases this channel has covered, a single person willing to look at a file that everyone else has stopped actively pursuing and decide that it is not actually finished. In 2023, Keelin Neikov, newly appointed as a detective with the New Haven Police Department, took on Andrea Reyes’s case. According to later reporting, Neikov reviewed the file and made a specific deliberate judgment call: she believed there was still solvability here, that despite the 23 years that had passed, there were still investigative avenues that had not been fully exhausted. This judgment reflects something important about how cold case investigation has changed over the past two decades. In 1999 and even in 2009, the tools available to trace a person who had disappeared into another country were extremely limited; investigators were dependent on formal law enforcement cooperation between countries, physical searches, and whatever leads happened to surface through traditional channels. By 2023, an entirely new set of investigative tools had become available and had matured into genuinely powerful instruments: social media platforms that most people maintain some presence on regardless of what country they live in, digital public records that can be cross-referenced in ways that were previously impossible, and forensic genetic genealogy techniques that had by this point cracked open dozens of cold cases across the United States that investigators had long since given up on solving through conventional means.

Detective Neikov went back through old interviews connected to the case—conversations conducted in the immediate aftermath of Andrea’s abduction in 1999 and potentially other contacts made in the years since. She pursued search warrants that gave her access to records and information that would not have been available or would not have been thought to search for in earlier years of the investigation. She turned to social media, a tool that by its very nature did not exist in any meaningful form when Andrea disappeared in 1999, but that by 2023 had become one of the single most powerful tools available to investigators trying to locate someone who believed quite reasonably that they had successfully disappeared decades earlier. Combining these threads, Neikov identified a woman living in Puebla, Mexico, whom she believed with real confidence was Andrea Reyes. Identifying a potential match and confirming it are two entirely different things, and the path between them required Detective Neikov to do something that could easily have gone wrong: reach out across an international border to a woman who had spent her entire remembered life believing a story about her own identity that may or may not have been true, and ask her to consider the possibility that everything she understood about herself was built on a lie. Neikov made contact with the woman she believed to be Andrea.

What was actually said in that initial conversation and how the woman initially responded to the suggestion that she might be a child who had been reported missing from Connecticut a quarter of a century earlier has not been fully detailed in public reporting. What is documented is the outcome: the woman agreed to submit a DNA sample. This is in itself a remarkable act of trust and courage. Agreeing to a DNA test in a case like this means opening yourself to the possibility that the family you have always known, potentially including the mother who raised you, is not exactly who you believe them to be, and that the life you have built is about to be examined by law enforcement in a country you may have no living memory of ever having lived in. Around this same period, the investigation’s momentum produced its own separate breakthrough: Andrea reached out to the man she believed to be her father. Whether this contact happened as a direct result of Detective Neikov’s outreach or developed somewhat independently as Andrea began processing the possibility that her identity was not what she had always understood it to be, the effect was the same. For the first time in 23 years, Carlos Reyes had a real direct connection to a woman who might be his daughter.

To move from suspicion and hope to certainty, New Haven police turned to Astrium, a Texas-based forensic laboratory that has become one of the most significant names in American cold case investigation over the past several years. Astrium specializes in forensic genetic genealogy, the same broad discipline of investigative science that has, across dozens of cases nationwide, identified decades-old unidentified remains, solved cold case murders, and, in cases exactly like this one, confirmed the true identities of individuals whose entire life stories had been built on false information. This case was processed as part of the lab’s project initiative specifically designed to help resolve long-term missing persons and unidentified persons cases by giving families certainty and giving law enforcement the evidentiary foundation needed to pursue justice. Kristen Middleman, the lab’s chief development officer, spoke to the broader significance of this kind of work when the case was announced: “Whether it is unidentified human remains, the identity of a suspect in a crime, or like in this case, confirming the identity of a woman who has been missing for decades, we do the same thing every day. We find answers for families.”

The lab conducted what is known as rapid relationship testing, comparing the DNA sample provided by the woman living in Puebla against a reference sample submitted by Carlos Reyes. This type of testing is specifically designed to establish or rule out close biological relationships, such as parent and child, with a high degree of statistical confidence without requiring the kind of extensive genealogical family tree construction that some other genetic genealogy cases demand when there is no existing candidate to test against. Because investigators already had a strong evidence-based candidate—Andrea herself—identified through Detective Neikov’s investigative work rather than through a blind database search, the testing process could move considerably faster than in cases where genealogists must build an entire family tree from scratch, cross-referencing distant cousin matches across multiple generations to eventually triangulate a single unknown identity. This distinction matters, and it says something important about how far this particular field of forensic science has advanced. In many of the cold cases genetic genealogy has resolved over the past several years identifying decades-old unidentified remains, investigators start with nothing but a DNA profile and no named suspect at all, requiring months of painstaking genealogical reconstruction before a single name can even be proposed. Andrea’s case moved faster precisely because good, patient, old-fashioned detective work—search warrants, social media investigation, the careful rereading of decades-old interview transcripts—had already narrowed the question down to a single specific person before the DNA lab ever became involved. The science confirmed what strong investigative instinct had already found.

That collaboration between traditional policing and modern forensic technology is increasingly the model that is cracking open cases that had been considered permanently unsolvable. The results confirmed a father-daughter relationship between Carlos Reyes and the woman living in Puebla, Mexico. She was Andrea Michelle Reyes, missing since October of 1999, found at last in early 2025—25 years, 4 months, and a handful of days after she was taken from a scheduled visitation in New Haven, Connecticut. Her identification marked the seventh case in Connecticut in which officials had used technology to confirm someone’s identity, part of a growing pattern across the country in which forensic genetic genealogy has become one of the single most effective tools available for resolving cases that traditional investigative methods had exhausted decades earlier. On March 12th, 2025, the New Haven Police Department held a press conference to formally announce Andrea Reyes’s identification and the resolution of a case that had haunted the department and one family for a quarter of a century. Carlos Reyes attended. It was a moment of joy, he said, describing the moment he learned with scientific certainty that his daughter was alive. Andrea’s stepmother, the woman who had, in the years since Andrea’s disappearance, become part of the family Carlos built while continuing to hold space for the daughter who was missing, delivered an emotional statement to the gathered press and cameras: “Thank you for sharing in our joy in finding our daughter, Andrea. After 25 years, God has answered our prayers and blessed us with a chance to know her again.” She continued with a candor that acknowledged the complexity ahead rather than pretending this was a simple, uncomplicated, happy ending: “We recognize that this reacquaintance will have challenges. However, we are confident that God is building the path before us and leading our steps.”

New Haven Police Chief Carl Jacobson addressed the broader significance of the case for his department’s approach to unsolved cases generally: “This case reflects the hard work of our officers and detectives. While cases may have investigative leads exhausted at the time, no cold case is ever truly closed. We remain committed to resolving every cold case, and this is a perfect example of that effort.” And Detective Keelin Neikov, the officer whose decision to revisit a 23-year-old file had ultimately cracked the case open, offered a characteristically measured, honest assessment of exactly what had been accomplished and exactly what remained ahead for the Reyes family: “I am happy that we could have this conclusion for us and bring them some resolve, but this is a whole new venture they’re about to go down.” As of the case’s public announcement in March of 2025, Andrea Reyes and her father had connected by phone and through online messaging—real direct contact after 25 years, but not yet the in-person reunion that both of them and the extended family waiting alongside them are working toward. The National Center for Missing and Exploited Children has been assisting the family in navigating what comes next. This detail matters, and it deserves attention rather than being treated as a footnote. Organizations like NCMEC do not simply confirm identities and close files; in cases involving this much time, this much distance, and this much fundamental disruption to a person’s sense of self, the actual work of reunification requires careful professional support, helping a 27-year-old woman who has spent her entire conscious life in Mexico, likely speaking primarily Spanish, culturally rooted in a life and a family structure she has always understood as fully her own, begin to build a relationship with a father, a stepmother, and potentially other biological family members who are in every practical sense strangers to her.

Andrea Reyes continues to live in Mexico. Her relationship with Rosa Tenorio, the mother who raised her, has not been publicly detailed—a deliberate and appropriate silence given how complicated that relationship almost certainly is. Whatever Rosa Tenorio did in 1999, and whatever she may or may not have told her daughter over the following 25 years about where they came from and why, she was also, in every everyday sense a child experiences motherhood, Andrea’s mother for the entirety of her life. Untangling that reality, the same psychological complexity this channel has encountered in several other cases involving long-term parental abduction, will likely be one of the most difficult parts of whatever comes next for Andrea. The felony warrant for Rosa Tenorio’s arrest, originally issued in 1999 and renewed in 2009, remains active today, but that warrant carries a significant and specific limitation: it is only valid within the United States. New Haven police have confirmed that they believe Rosa Tenorio still resides in Mexico, and because she is Andrea’s biological mother, Mexican authorities have made clear that the country will not extradite her to face the custodial interference charge pending against her in Connecticut.

This is the same fundamental legal reality that appears to explain why Mexico declined to intervene when authorities located Rosa Tenorio and Andrea back in the year 2000: a biological parent, even one who has violated a foreign court’s custody order and evaded an active arrest warrant for a quarter of a century, occupies a different legal category than a stranger or non-relative who commits an equivalent act. Many countries, Mexico included, are simply far less willing to forcibly remove a child from a biological parent or to extradite that parent for prosecution abroad, regardless of the legal violations involved under the laws of the country the child was taken from. For Carlos Reyes, this means that even now, even with his daughter identified, even with direct contact reestablished, even with a press conference and a DNA report and 25 years of searching finally answered, the woman who took his daughter from him during a routine visitation in 1999 will very likely never face any legal consequence for what she did. The case of Andrea Reyes sits at the intersection of two of the most persistent and difficult challenges in American missing children cases: the limitations of cross-border law enforcement cooperation and the transformative power of modern forensic genetic genealogy to finally answer questions that decades of conventional investigation could not.

The systemic failure at the heart of this case is not subtle. Mexican authorities located Rosa Tenorio and Andrea within roughly a year of the abduction and made a decision at some level of government not to intervene. Whatever legal, diplomatic, or bureaucratic reasoning produced that decision, the practical result was that a 2-year-old girl taken in direct violation of a valid American custody order was left in the exact place where she had already been found for the next 22 years, while her father continued searching for a needle that authorities had at one point already located precisely. This case is not unique in that respect. International parental abduction cases involving a biological parent occupy a genuinely difficult legal space, one where the enforcement mechanisms that exist on paper—warrants, custody orders, extradition treaties—frequently collide with a receiving country’s reluctance to forcibly separate a child from a parent who, whatever their legal violations under foreign law, is still that child’s biological mother or father. The Hague Convention on the Civil Aspects of International Child Abduction exists specifically to try to address this gap, establishing a framework for the return of children wrongfully removed across international borders, but its effectiveness depends entirely on both countries being signatories and on the specific circumstances of each case, and it has never been a complete solution to the problem this case illustrates so starkly.

And then there is the other half of this story. The technology that finally cracked the case open a full 2 years after conventional police work alone had been exhausted for over two decades. Forensic genetic genealogy, the discipline that companies like Astrium specialize in, has fundamentally changed what is possible in cold case investigation.

A case that sat dormant, technically open, but practically unsolvable for 23 years, was resolved within roughly two years of a single detective deciding to apply modern tools to an old file. Social media searches, updated public records access, and rapid DNA relationship testing accomplished in a relatively short window what an entire apparatus of international law enforcement cooperation had not managed to accomplish in over two decades.

That contrast between what conventional cross-border enforcement could not do and what modern genetic identification technology eventually did do is perhaps the single most important lesson this case offers. No cold case, as Chief Jacobson said, is ever truly closed. But it often takes exactly the kind of technology this case relied on to prove that statement true rather than merely aspirational.

There is one detail from this case that I have not been able to stop thinking about since I first read it. In the year 2000, Mexican authorities knew exactly where Andrea Reyes was. They found her, and somewhere in a government file, a decision was made that no further action would be taken to return her to the father who had legal custody of her and who was at that very moment desperately searching for her.

For the next 22 years, Carlos Reyes continued that search: trips to Mexico, cooperation with the FBI, and watching age-progression images render a daughter he could not locate, while the answer to where she actually was sat unused in a file that had already been closed. It took a newly appointed detective deciding in 2023 that a 23-year-old case was not actually finished. It took social media, search warrants, and a rapid DNA test from a lab in Texas. And it took a woman in Puebla, Mexico—a woman who had built an entire adult life under one understanding of who she was—agreeing to have that understanding tested against the truth.

What do you think happened in the year 2000 when Mexican authorities found Andrea and chose not to act? Do you think justice will ever be possible for Rosa Tenorio given that she can likely never be extradited? Drop your thoughts in the comments below and tell me where in the world you are watching from.

Andrea Michelle Reyes, taken during a scheduled visitation in New Haven, Connecticut in October of 1999 at 23 months old, found in Puebla, Mexico, and confirmed by DNA in March of 2025, 25 years later. Her father never stopped searching, and somewhere for 22 of those years, the answer was already known.