Children's rights · 12 min read

What children of different ages actually understand about separation

Infants do not understand divorce. Adolescents understand it too well. A research-grounded look at how children at each developmental stage process a parental separation, what helps, and what causes lasting harm.

← Back to articles

One of the most consequential mistakes parents make during a separation is assuming their child understands what is happening the way an adult does. A three-year-old does not understand divorce. A fifteen-year-old often understands it far more completely than the parents realize. Treating both as if they need the same conversation, the same reassurance, or the same level of information produces predictable harm at both ends.

The research on how children of different developmental ages experience a parental separation is substantial, long-running, and surprisingly consistent. It also pushes back firmly against several beliefs parents tend to hold about their own children. Before going stage by stage, two findings from the literature apply across every age and need to be stated plainly.

First, it is the conflict, not the divorce itself, that drives most of the documented harm to children. A 2022 review in Clinical Child and Family Psychology Review synthesizing decades of research concluded that the pre- and post-divorce family environment, primarily the level of interparental conflict, accounts for most of the variability in children's adjustment, not the legal divorce per se (Cao, Fine and Zhou, 2022). The American Psychological Association states it even more directly: children often do worse when parents remain in high-conflict marriages than when those marriages end.

Second, divorce is a process, not a single event. Researchers now frame it as a multi-year unfolding, beginning before the legal separation and extending for years afterward. What happens during that process, especially what children are exposed to, shapes their trajectory far more than the legal status change on any single date.

Infants and toddlers (0 to 3)

Children in this age range have no conceptual framework for divorce. They cannot grasp legal status, adult relationships, or future continuity. What they can perceive, and perceive with full force, is the disruption of their primary attachment relationships and the emotional climate of the people caring for them.

Attachment research is the foundation here. Infants organize their entire sense of safety around one or a small number of primary caregivers. Any significant change in access to those caregivers registers as a threat, not as a family restructuring. The AFCC think tank on overnight care of young children, led by Marsha Kline Pruett, Jennifer McIntosh and Joan Kelly (2014), noted that for infants, parental separation is essentially the experience of losing regular access to one caregiver, because young children lack the object permanence and temporal reasoning needed to understand that an absent parent still exists and will return.

What this looks like in behavior: heightened separation anxiety at transitions, sleep disturbance, regression in toilet training and feeding, irritability that has no obvious source, clinginess to the primary caregiver. A peer-reviewed study in BMC Pediatrics (Kacenelenbogen et al., 2016) found parental separation associated with psychomotor developmental risks in children aged 28 to 32 months, particularly in language and social-emotional domains, with the effect mediated by parental stress and reduced quality of interaction.

What helps

  • Predictable, consistent routines in both households. Same bedtime ritual, same feeding rhythm, familiar objects that travel with the child.
  • Warm, attuned parenting from both caregivers. A parent's own emotional regulation is the single most important buffer at this age.
  • Frequent, shorter contact rather than long absences. Pruett, McIntosh and Kelly recommend that for infants under 12 months, long overnight separations from a primary caregiver be phased in carefully based on the child's attachment security.
  • Minimizing the child's exposure to adult conflict at handoffs. Raised voices and tense exchanges are directly physiologically stressful to an infant, even when no words register.

What harms

  • Long gaps between contact with either parent, especially the primary attachment figure.
  • Parental emotional unavailability. A severely depressed or preoccupied parent creates a second attachment disruption on top of the first.
  • Hostile exchanges at transitions.
  • Abrupt, unplanned changes in caregiving arrangements.

What to say

Infants and toddlers do not understand verbal explanations of divorce, but parents still need to speak to them. Not to explain the adult situation, but to provide emotional scaffolding. Zero to Three advises naming the child's feelings and reminding them of the routine: "I know it's hard to say goodbye. We will miss each other. You and Daddy are going to have a great time together, and I will see you again in two days." Concrete, present, reassuring. Photos of both parents displayed in each home help toddlers maintain an internal sense of an absent caregiver. Do not attempt to explain reasons.

Preschool children (3 to 5)

Preschoolers operate in what Piaget called the preoperational stage: egocentric, magical, non-logical. They understand the fact of a parent leaving but cannot grasp causation, the permanence of adult relationships, or the irreversibility of divorce. Their cognitive egocentrism leads directly to self-blame. If something bad has happened, it must be because of something they did.

Judith Wallerstein's foundational longitudinal work documented this with painful clarity. Preschool-aged children were the most likely of any age group to believe they had caused the divorce. Through a tantrum, through a wished-for thought that a parent would "go away," through a magical logic that their actions had consequences they could not understand. In her 25-year follow-up (Wallerstein and Lewis, 2004, Psychoanalytic Psychology), feelings of loneliness, bewilderment and anger were especially powerful, decades later, among those who had been six or younger at the time of the breakup. They had the least capacity to comfort themselves.

The American Academy of Child and Adolescent Psychiatry states it plainly: children often believe they have caused the conflict between their parents, and many take on the responsibility for trying to bring them back together, which only adds to their stress.

What this looks like

  • Magical thinking about reconciliation, sometimes lasting for years.
  • Regression to earlier behaviors: bedwetting, baby talk, clinging, separation anxiety that had previously resolved.
  • Grief and sadness expressed through play more than words.
  • Fear of being abandoned by the remaining parent.
  • Increased aggression or tantrums as a behavioral proxy for feelings too big to name.
  • Somatic complaints with no medical cause: stomachaches, headaches, sleep problems.
  • Confusion about logistics. Where will I sleep? Who is picking me up?

What helps

  • Explicit, repeated reassurance that the child did not cause the divorce and cannot fix it. A child who is not told this directly will almost universally conclude that they did and they can.
  • Consistent contact with both parents. A 2017 study in Acta Paediatrica (Sundelin Wahlsten et al.) found preschoolers in joint physical custody arrangements showed significantly fewer psychological symptoms than those living primarily or exclusively with one parent.
  • Simple, age-appropriate language and the patience to answer the same question many times.
  • Predictable schedules in both homes. At this age, routine is the scaffolding that holds emotional regulation together.
  • Coparenting quality. A 2021 study in BMJ Paediatric Open(Bergström et al.) of nearly 13,000 Swedish three-year-olds found that once coparenting quality was accounted for, children in different living arrangements showed similar mental health. Coparenting quality was a stronger predictor than the specific custody arrangement.

What harms

  • Failure to explicitly address self-blame.
  • Treating a four-year-old as a source of emotional comfort or information.
  • Badmouthing the other parent. Preschoolers internalize both parents as part of their own identity. Attacking one is experienced by the child as an attack on themselves.
  • Introducing new romantic partners quickly.
  • Broken promises about contact.

What to say

Keep explanations brief, concrete, and repeated. The AACAP advises something close to: "Mommy and Daddy have decided not to live together anymore. That sometimes happens with grown-ups. It is not because of anything you did. We both love you and will always be your mom and dad." Be ready to revisit the conversation many times over many months. Each repeated question is an opportunity for reassurance, not a sign that the earlier explanation failed.

School-age children (6 to 12)

School-age children have reached concrete operational thinking. They understand cause and effect, can hold multiple perspectives, and grasp the permanence of divorce. This cognitive advance is, paradoxically, a source of additional pain. They now understand that the divorce is real, lasting, and not their fault, but also that they cannot fix it.

Wallerstein's team documented a characteristic school-age response: a shift from self-blame toward anger, grief, and profound loyalty conflict. These children are old enough to understand that each parent is a separate person with claims on their loyalty, and they often experience the divorce as placing them in an impossible position. They observe their parents carefully and form strong moral judgments about each.

The American Academy of Pediatrics' 2016 clinical report noted that many children show short-term, painful feelings and bounce back within two to three years, but that divorce can also be associated with longer-term academic, behavioral, social and emotional problems.

What this looks like

  • Academic decline. A consistent finding across the research. Attention and working memory are depleted by ongoing family stress.
  • Loyalty conflicts. Enjoying time with one parent can feel like betraying the other. Children become secretive about what happens at the other house.
  • Intense, conscious grief. Unlike preschoolers, they know what they have lost and mourn it.
  • Anger, often directed at the parent perceived as having caused the divorce.
  • Shame and social withdrawal. Fear of being different from peers.
  • Psychosomatic symptoms.
  • Parentification. Children at this age are developmentally capable of providing real emotional support to a distressed parent, which many divorced parents inadvertently invite.

Research by Mazur et al. (1999, Child Development) found that negative cognitive errors about divorce events, including catastrophizing and self-blaming appraisals, predicted worse psychological adjustment. Children who could appraise the divorce situation more accurately fared significantly better. This has a practical implication: correcting distortions directly is therapeutically effective at this age.

What helps

  • Honest explanation at a level that respects their cognitive capacity. They can handle more than preschoolers, but they still do not need adult details about fault, finances, or infidelity.
  • Explicit, repeated permission to love both parents without guilt. "It is okay to love Dad. I want you to have a good time with him."
  • Stable routines at school and in extracurriculars. The structure of school is a particularly important buffer at this age.
  • Authoritative parenting in both homes. Hetherington's Virginia Longitudinal Study found parenting quality was the strongest within-family predictor of adjustment.
  • Financial stability where possible. Loss of activities, neighborhood changes, and school disruption are major secondary stressors.

What harms

  • Using the child as a messenger, spy, or confidant. "What did Dad say about me?"
  • Conflict in the child's presence. The physiological evidence on this is overwhelming.
  • Undermining the other parent's authority in a way that is weaponized rather than simply different.
  • Pushing the child to choose sides. The AACAP is explicit: long custody disputes or pressure to take sides can be particularly harmful.

Adolescents (13 to 18)

Adolescents have reached formal operational thinking. They can reason abstractly, hold multiple perspectives simultaneously, understand the adult relational dynamics at play, and consider long-term implications for their own lives. They fully understand what divorce means. They often understand more about their parents' marriage than the parents realize.

This maturity creates a distinctive set of challenges. Adolescents are in the process of forming their own identity partly by evaluating their parents, sometimes by repudiating them. Divorce can accelerate that process in destabilizing ways. They can and do worry, explicitly, about whether they will be able to maintain lasting romantic relationships of their own. Wallerstein called this the "sleeper effect": problems that were not apparent in childhood crystallized in adolescence and young adulthood as teens began forming romantic attachments of their own.

The TRAILS longitudinal cohort study in the Netherlands, following 2,230 adolescents from age 10 to 23 (Xerxa et al., 2021, European Child and Adolescent Psychiatry), is the most methodologically rigorous recent study on adolescent timing. It found that levels of both internalizing and externalizing problems were significantly higher in the period after parental divorce, but not in the period before, with a persistent and increasing effect over four years of follow-up compared to adolescents who did not experience a divorce. This directly refuted the comforting idea that adolescent problems mostly preceded the divorce.

What this looks like

  • Gender patterns documented across TRAILS, Hetherington's work, and others. Female adolescents tend more toward internalizing problems (depression, anxiety, over-controlled behavior). Male adolescents tend more toward externalizing problems (acting out, conduct issues, substance use). These are group tendencies with significant individual variation.
  • Disengagement from family. Teens are separating from parents anyway, and divorce can accelerate premature pseudo-independence.
  • Parentification. Wallerstein documented this extensively. Older children in divorced families often took on substantial caretaking of younger siblings and emotional caretaking of needy parents. The price, for those who did too much over too many years, was the loss of their own adolescence.
  • Earlier and heavier substance use, earlier sexual activity, on average.
  • Academic disruption.
  • Anger and a strong sense of injustice. Teens are developmentally tuned to fairness, and they often view the divorce as something done to them.

Hetherington's broader finding remains important context. About 20 to 25 percent of children of divorce showed clinical-level problems in adulthood, compared with about 10 percent in intact families. A real difference, but also a clear statement that the majority did not develop those problems. Divorce raises risk. It does not determine outcome.

What helps

  • Respect for their cognitive maturity. They do not need to be shielded from adult complexity, but they do need to be protected from adult burden.
  • A clear distinction between parental and adolescent roles. Get adult emotional support from adults.
  • Continued authoritative parenting in both homes. Consistent rules, warmth, and monitoring. Hetherington found this was the single strongest protective factor for adolescent outcomes.
  • Acknowledgment of their feelings, including their anger.
  • Financial planning for their education. Wallerstein found that only about 30 percent of children of divorce in her sample received consistent college financial support, compared with about 90 percent in her intact-family comparison group. This is a concrete, fixable harm.

What harms

  • Treating them as peers or confidants. The most corrosive dynamic in Wallerstein's qualitative data.
  • Putting them in the middle of financial or custody disputes.
  • Rigidly enforcing court-ordered visitation that overrides the teen's own stated, reasoned preferences. Wallerstein found this generated lasting anger.
  • Sustained conflict that continues for years post-separation. The TRAILS data showed problems getting worse, not better, over four years.
  • Hastily introduced new romantic partners.

A few common assumptions worth letting go of

Young children do not "just bounce back" because they do not understand. They understand attachment disruption, even if they have no language for it. Wallerstein's strongest qualitative finding was that the children youngest at the breakup carried the most powerful adult memories of loneliness.

Keeping kids "out of it" by telling them nothing makes things worse, not better. Uncertainty is more frightening than age-appropriate information, and children fill in gaps with their own worst guesses. The AACAP is explicit on this.

Teenagers are not "old enough to handle it" in a way that means they need less support. The TRAILS data suggest they sometimes need more. They need different support, not less of it.

And both parents being present and involved is not, by itself, enough. Wallerstein found that remaining in frequent contact with both parents did not, on its own, alleviate adult suffering. Quality of relationship and minimal conflict between the parents mattered more than the simple fact of contact.

None of this is a counsel of despair. Hetherington's resilience finding is real, and the protective factors are well established: low conflict between parents, warm and consistent parenting, economic stability, school connectedness, and time. What the research is consistent about is that the adults around the child have substantial influence over which trajectory that child ends up on. That is a burden. It is also a reason for some quiet, sober hope.

Published May 2026.

Read next