The purpose of this study was to evaluate the accuracy of clinical diagnosis of acute pelvic inflammatory disease (PID). Data were obtained on 176 consecutive women admitted to St. Elizabeth Hospital Medical Center with a clinical diagnosis of PID. All underwent diagnostic laparoscopy. PID was established laparoscopically in 134 (76.1%) of the patients. Statistical tests for significant associations between PID and each of 21 clinical indicators of the disease were conducted using the chi 2 and Mann-Whitney tests. Stepwise logistic regression was performed on those variables whose univariate tests of significant association with PID resulted in P values < 0.20. An optimal set of PID indicators consisted of adnexal tenderness, lower abdominal pain of < one week's duration and an elevated white blood cell count. Use of these indicators resulted in a test with an estimated sensitivity and specificity of 86.6% and 45.7%, respectively. Estimated predictive values for positive and negative test results were 0.84 and 0.52, respectively. These results confirm the fact that laparoscopy is the definitive diagnostic modality in PID.