We examined the utility of patient self-report forms in identifying those gynecologic oncology patients who would be diagnosed by an experienced consultation-liaison psychiatrist as suffering from major depression. Sixty-five women with gynecologic tumors were evaluated by a consultation-liaison psychiatrist, using standardized (DSM-III) criteria. Each patient also completed a Carroll Rating Scale for Depression (CRS). The CRS demonstrated sensitivity and specificity of 87 percent and 58 percent, respectively. Used as a screening instrument to rule out depression, the CRS yielded a negative predictive value of 94 percent. We identified a priori forty items from the CRS which should not be influenced by the non-psychiatric biologic effects of gynecologic tumors, and compared the performance of this non-cancer related symptoms subscale (NCSG) to that of the CRS. The NCSG did not significantly outperform the CRS; its sensitivity and specificity were 87 percent and 62 percent, respectively. Because our study population was relatively homogeneous (i.e., non-ovarian gynecologic oncology patients without severe debilitation who were not receiving chemotherapy, radiation therapy, or other invasive procedures), the findings should not be generalized to other oncologic populations at this time. Our results suggest that patient self-report forms can be effective screening devices for identifying those non-ovarian, gynecologic oncology patients who should then be carefully evaluated for coexisting clinical depression.