How a custody evaluation actually works

← Back to articles

A custody evaluation is, for many separated parents, the most consequential professional assessment they will ever undergo. The report that emerges, often 30 to 100 pages long, is typically the single most influential document a judge will read on the case. Most parents arrive at the process knowing almost nothing about how it works. This article is a plain-language guide to how custody evaluations are ordered, who conducts them, what they involve, how courts use them, and the documented limits parents should understand going in.

When and why a court orders one

Custody evaluations are not routine. The American Psychological Association's 2010 Guidelines for Child Custody Evaluations in Family Law Proceedings note that roughly 90 percent of parents reach a custody arrangement on their own. Evaluations arise in the remaining 10 percent of cases, typically when one or more of the following conditions are present: serious allegations of child abuse or neglect, domestic violence allegations, disputed claims about mental health or substance use, allegations of parental alienation, intractable high-conflict litigation, relocation disputes, or contested motions to modify an existing order.

State court rules codify the authority to order evaluations. California Rule of Court 5.220(b) authorizes evaluations to assist the court in determining the best interests of the child, focusing on health, safety, welfare and best interests. Texas Family Code section 107.103 authorizes the court, on its own motion or on the motion of a party, to order a child custody evaluation in any proceeding in which the court has jurisdiction. Most states have analogous authority.

Who conducts them

The two governing national frameworks are the APA's 2010 Guidelines (with a 2022 revision now in effect) and the Association of Family and Conciliation Courts' 2006 Model Standards of Practice for Child Custody Evaluation, updated in 2022 as the Guidelines for Parenting Plan Evaluations in Family Law Cases. The AFCC standards apply to any mental health professional offering custody recommendations, not just psychologists. The APA standards apply specifically to psychologists.

The AFCC's Standard 1.2 sets a floor: evaluators must hold at least a master's degree in a mental health field, with formal training in child development, child and adult psychopathology, interviewing techniques, and family systems. State licensing rules typically add more. California (Family Code section 3110.5 and Rule of Court 5.225) requires that court-connected or private evaluators complete a domestic violence and child abuse training program. Colorado, under Chief Justice Directive 21-02 (amended 2024), requires that Parental Responsibility Evaluators be licensed Colorado mental health professionals registered with the Office of the Child's Representative.

In practice, evaluators are most commonly licensed psychologists (PhD or PsyD), because psychologists are typically the only licensed mental health professionals authorized to administer and interpret objective psychological testing. Licensed clinical social workers and licensed professional counselors conduct more limited evaluations in some jurisdictions. Court clinic staff serve in jurisdictions with family court services offices. Psychiatrists are used less frequently, often for complex medication or serious mental illness questions.

Both governing frameworks insist that evaluators function as impartial examiners, not as advocates for either party. The AFCC's Preamble is blunt on this: regardless of how arrangements were made and regardless of who is paying, the evaluator "shall always function as an impartial examiner." Even if you pay the evaluator's fees, the evaluator's client is the court, not you.

What an evaluation actually involves

A well-conducted evaluation typically takes two to six months, sometimes longer in complex cases, and produces a written report that may run anywhere from 30 to over 100 pages. Both the APA Guidelines and the AFCC standards require a multi-method, multi-source approach. AFCC Standard 5.4 requires that evaluators use multiple data-gathering methods in order to increase accuracy and objectivity.

Clinical interviews with each parent

Each parent receives multiple individual interviews, typically three to five, conducted separately. These cover developmental and relationship history, perception of the child's needs, parenting practices and discipline style, coparenting willingness and attitude toward the other parent's role, allegations against the other parent, and prior mental health treatment, substance use history, and domestic violence history. The interviews are forensic assessments, not therapy. The evaluator is not your counselor, and everything you say can appear in the report.

Observation of parent-child interaction

Structured or semi-structured observations of each parent with each child, usually in the evaluator's office and sometimes in the home. AFCC Standard 10.2 requires that observations be of sufficient duration and frequency, and that each parent be observed with each child. The standards also note that evaluators should be aware that their presence affects behavior. Both parents and children tend to present their best selves under observation, and evaluators are trained to read the data alongside other sources.

Child interviews

Children are interviewed separately from parents. The manner and number of interviews depend on the child's age and developmental level. AFCC Standard 9.1 stresses non-leading, open-ended questioning. The child's expressed preferences are one factor, not the determinative one. Most evaluators are trained in forensic child interviewing protocols (such as the NICHD Protocol or RATAC), particularly when abuse allegations are involved.

Psychological testing

This is the component most often scrutinized in court. AFCC Standard 6.1 requires evaluators to consider whether to use formal assessment instruments based on the extent to which an instrument directly measures the psychological constructs at issue.

The most commonly used tests with parents include the MMPI-3 (Minnesota Multiphasic Personality Inventory, 3rd edition, published 2020), a 335-item self-report objective personality inventory with ten validity scales that are specifically designed to detect over- and under- reporting; the MCMI-IV (Millon Clinical Multiaxial Inventory, 4th edition) for personality disorders; the PAI (Personality Assessment Inventory); the Rorschach Inkblot Technique (whose validity in custody contexts is debated); the PSI-4 (Parenting Stress Index); and occasionally the WAIS-IV for cognitive functioning. With children, tests may include the WISC-V, projective measures like the Children's Apperception Test, the Trauma Symptom Checklist for Children, and the Child Behavior Checklist (parent, teacher, and youth report).

Survey data from Otto, Edens and Barcus (2000, Family and Conciliation Courts Review) found that 92 percent of evaluators used the MMPI/MMPI-2 in 91 percent of their evaluations, making it by far the dominant adult instrument. Mulchay (2022, Journal of the American Academy of Matrimonial Lawyers) provides a current family-law primer on the MMPI-3, which has updated norms and ten validity scales engineered to flag a parent who is trying to look better than they are. The validity scales are one of the reasons that attempting to "perform" for a custody evaluator is not a strategy.

A critical limitation, originally noted by Grisso in 1984 and reiterated since, is that standard psychological tests were not designed to assess parents' relationships with children or to predict parenting capacity directly. Inferring parenting capacity from a measure of general psychopathology is a leap, and one that can be challenged in court.

Collateral contacts

AFCC Standard 11.1 calls collateral information essential. The importance of collateral source information cannot be overestimated. Interview data from parents must be supplemented and tested against third-party sources. Typical collaterals include teachers and school counselors, pediatricians and treating physicians, therapists, family members and friends, law enforcement records, child protective services records, prior evaluators or treating experts, and daycare providers or coaches. Standard 11.3 requires that evaluators corroborate collateral information rather than simply parrot it, but third-party statements can be powerful.

Home visits

Not universally required, but commonly conducted in evaluations involving young children, neglect allegations, concerns about living conditions, or disputes over parental living arrangements. The evaluator observes physical safety, the adequacy of the home environment, age-appropriate space for the child, parent-child interaction in the natural setting, and the presence of significant others.

Document review

Court pleadings, prior custody orders and declarations, medical and school records, prior mental health treatment records, communications (text messages, emails, increasingly social media), police reports, CPS records, and prior evaluation reports.

The report and how courts use it

The evaluation culminates in a written report submitted to the court and provided to the parties and their attorneys. The report typically includes identifying information and reason for referral, the methods and sources used, a summary of each party's self-report and history, behavioral observations from interactions and child interviews, psychological testing results interpreted in forensic context, integration of collateral information, the relevant legal factors under the applicable state statute, clinical opinions, and (often, though not always) recommendations.

APA Guideline 13 states that psychologists, where they make recommendations at all, should base them on the psychological best interests of the child and should be mindful of the limits of their ability to predict future behavior. Whether to make recommendations at all is within the evaluator's discretion. When made, recommendations typically address legal custody (sole or joint decision-making authority), the residential schedule, therapeutic referrals, and any conditions or restrictions (supervised visitation, drug testing, reunification protocols).

The report is not binding on the court. The court retains decision- making authority. In practice, however, judges give custody evaluation reports substantial weight in contested cases, particularly when the evaluator testifies and the report has survived cross-examination.

The documented limits

The most influential scholarly critique of custody evaluations is by attorney Timothy Tippins and psychologist Jeffrey Wittmann, "Empirical and Ethical Problems with Custody Recommendations: A Call for Clinical Humility and Judicial Vigilance," 43 Family Court Review 193 (2005). Their core arguments are worth understanding because they shape how evaluations are properly challenged.

There is no validated, peer-reviewed body of research demonstrating that mental health professionals can reliably predict which parenting arrangement will be in a particular child's best interests.
"Best interests" is a legal and moral standard, not a clinical one. The ultimate determination involves value judgments that belong to judges, not to mental health experts. Evaluators who make ultimate- issue recommendations risk crossing into judicial territory.
Standard psychopathology measures (the MMPI, MCMI, PAI) were designed to assess clinical constructs like depression and anxiety, not parenting behavior or child outcomes. Inferring one from the other is scientifically uncertain.
Evaluators sometimes overstate the strength of their methods. Tippins and Wittmann called for "clinical humility": explicit acknowledgment of the limits of the data.

Emery, Otto and O'Donohue's 2005 article in Psychological Science in the Public Interest, "A Critical Assessment of Child Custody Evaluations: Limited Science and a Flawed System," made parallel arguments and documented additional structural problems: inconsistent methodology across evaluators, no feedback loop on whether recommendations served children well, equity disparities from the $5,000-to-$30,000-plus cost of private evaluations, role confusion between treating therapists and forensic evaluators (which the APA and AFCC both prohibit), and bias along gender, ethnicity and socioeconomic lines. None of this means evaluations are without value. It means they are evidence to be weighed, not oracles to be deferred to.

What parents should know going in

Eight points that matter in practice.

This is a forensic investigation, not therapy. Anything you say, any document you provide, any behavior observed can appear in the report. APA Guideline 9 requires that informed consent cover the potential uses of the data, including disclosure in litigation.

You will be tested, not just interviewed. The MMPI-3's ten validity scales are specifically designed to detect under-reporting of symptoms. Trying to "look good" is more likely to be flagged than to succeed.

Cooperation is mandatory and resistance is noted.Courts can draw adverse inferences from a parent's refusal to participate. AFCC Standard 5.5 requires a balanced process. APA Guideline 10 addresses the impact of non-cooperation on the validity of the evaluation.

Collateral contacts are beyond your control.Evaluators will speak with teachers, doctors, therapists, daycare providers, sometimes family members. You cannot control what these people say. AFCC Standard 11.3 requires that the evaluator corroborate, but the statements still enter the record.

The evaluator is the court's expert, not yours. Even when you are paying. Attempts to cultivate the evaluator as an ally are ethically prohibited on the evaluator's side and likely to backfire on yours.

Recommendations are not binding, but they carry weight.Judges follow them in most contested cases, particularly when the evaluator testifies and the report survives cross-examination.

You can challenge the evaluation. Parents have the right to depose the evaluator, retain a reviewing expert (sometimes called a forensic consultant) who critiques the methodology without conducting their own evaluation, cross-examine the evaluator at trial on qualifications, methodology, test interpretation, and compliance with APA and AFCC standards, and, under Daubert or Frye standards, challenge admissibility of specific test evidence. Jonathan Gould and David Martindale's The Art and Science of Child Custody Evaluations (Guilford Press, 2009) is the scholarly benchmark for methodologically rigorous practice and is commonly used by reviewing experts.

The evaluation has limits the evaluator must acknowledge.Under APA Guideline 13, psychologists are required to be mindful of the limits of their data and the limits of their ability to predict future behavior. AFCC Standard 12.4 requires that limits be articulated in the final report. An evaluator who presents recommendations with unqualified certainty is presenting work that may itself be open to challenge.

A custody evaluation is, at its best, a careful, multi-source attempt to give a court something more than competing affidavits to work with. It is also, at its honest, a process operating well past the edge of what the science can reliably support. Both things are true. Understanding them lets a parent prepare for the process without being either intimidated by it or naive about what it can deliver.

Published April 2026.