Statistically significant reductions in scores from initial pretreatment testing to subsequent pretreatment testing in self-report measures are a widespread phenomenon (Arrindell, 2001). If valid, these reductions reflect improvement in psychological functioning absent any formal treatment (Arrindell, 2001). Many explanations for these reductions in scores have been offered (Arrindell, 2001) including assessment reactivity (Epstein et al., 2005), social desirability (Henderson, Byrne, & Duncan-Jones, 1981); and mechanical responding (Bromet, Dunn, Connell, Dew, & Schulberg, 1986; Durham et al., 2002). However, relatively few studies have examined this phenomenon empirically. Even fewer studies have examined pretreatment improvements in parent ratings. The present study sought to replicate the pre-treatment effect and to evaluate potential causes of these changes, specifically, repeated assessment, social desirability, and mechanical responding. Additionally, the study examined the relationship between maternal depression and parent ratings. Participants included 28 self-referred mothers and their children, though only 17 completed the study in its entirety. Results indicated pretreatment reductions in maternal ratings of child symptomatology and parenting stress. Repeated assessment was not supported as a cause of pretreatment improvements (score reductions). Social desirability, however, may have been a factor in the mothers’ ratings. Mechanical responding did not appear as a likely cause. Maternal symptoms of depression were associated with less change over the course of the study. Clinical implications, study limitations, and suggestions for future research are discussed.