Abstract: Objective To investigate the correlation between female obsessive-compulsive disorder (OCD) and estradiol (E2) and prolactin (PRL). Methods A total of 90 female OCD patients who were treated in Yangzhou Wutaishan Hospital of Jiangsu Province from September 2020 to April 2023 were selected as subjects (an observation group). Meanwhile, another 90 healthy women without OCD who underwent physical examination were chosen as a control group. Both groups were compared for serum E2 and PRL levels. Receiver operating characteristic (ROC) curves were plotted to evaluate the diagnostic efficiency of serum E2 and PRL for OCD. Pearson correlation analysis was conducted to investigate the correlation between E2 and PRL levels and Yale-Brown Obsessive Compulsive Scale (Y-BOCS) scores. After treatment for three months, the observation group was subdivided into a good prognosis group (n=68) and a poor prognosis group (n=22) based on Y-BOCS score reduction. Both subgroups were compared for serum E2 and PRL levels. ROC curve analysis was performed to evaluate the value of serum E2 and PRL in predicting the prognosis of OCD patients. Results The observation group showed lower E2 and PRL levels than the control group (P<0.05). The area under the curve (AUC) values for E2, PRL, and their combination in diagnosing OCD were 0.840, 0.798, and 0.909, respectively, indicating good diagnostic value (P<0.05). The combined diagnostic efficiency was higher than that of E2 or PRL alone (z=3.314 and 4.095, P<0.05). Furthermore, Y-BOCS scores were significantly negatively correlated with both E2 and PRL (P<0.001). The good prognosis group presented higher E2 and PRL levels than the poor prognosis group (P<0.05). The AUC values for E2, PRL, and their combination in predicting the prognosis of female OCD were 0.826, 0.732, and 0.872, respectively, with good diagnostic value (P<0.05). The combined diagnostic efficiency was not significantly different from E2 alone, but superior to PRL alone (z=2.677, P=0.007). Conclusions E2 and PRL are significantly negatively correlated with female OCD, which can effectively diagnose OCD and predict patient prognosis. These indicators may be useful for the early clinical diagnosis and prognosis evaluation of OCD.