One variable that predicted the use of a social scientist as an expert witness was the nature of the alleged offense. Expert testimony was more likely to be involved when the alleged sexual abuse was intrusive, more frequent, and when the alleged perpetrator was a caregiver for the child. Another variable related to the use of expert testimony was the complainant's age when the abuse began. Cases in which the complainants were between the ages of 7 and 9 years old when the abuse began were less likely to involve expert testimony than cases in which complainants were between 13 and 19 years old when the abuse began. Also, expert testimony was significantly more likely to be used in cases that involved claims of repression as the reason for the delay in bringing the allegation. Expert testimony was more likely when complainants were female, perhaps because these cases more often involved the claim of repression. Seven variables predicted the presence of an expert at sentencing: the nature of the offense, frequency of abuse, length of delay to trial, the presence of threat, trial date, plea, and age difference between the complainant and the accused. Quicklaw was used to identify 2,064 criminal cases brought in Canada that involved delayed allegations of child sexual abuse. The cases were coded on a variety of variables that were used to predict the presence of an expert at trial and in the assessment of convicted offenders for sentencing. 2 tables, 60 references, and appended examples of experts' statements and definitions of categorical variables