The authors compared 134 6- to 12-year-old children from a military psychiatric clinic with a control sample to determine the salience of various risk factors in predicting levels of child psychopathology. Parents provided demographic information and completed standardized questionnaires on themselves and their children, while children completed two self-report symptom inventories. Results indicated that all hypothesized risk factors mediated effects on child psychopathology, but the effects of various risk factors differed as a function of the rater and type of psychopathological construct being measured. Generally, parental psychopathology and life stress mediated the greatest effects on overall child symptoms levels. Furthermore, the clinical and community samples differed in the presence and extent of risk factors. Results indicate the need for caution in studies of child psychopathology using only clinical samples and may suggest the importance of therapies based on environmental manipulations for a substantial proportion of patients seeking child psychiatric assistance.