Objective: To evaluate the ability of serum anti-M$ddot{u}$llerian hormone (AMH), FSH, and age to clinically predict ovarian response to controlled ovarian hyperstimulation (COH) in IVF patients with endometriosis. Methods: We evaluated 91 COH cycles, including 43 cycles with endometriosis (group I) and 48 cycles with male factor infertility (group II) from January to December, 2010. Patients were classified into study groups based on their surgical history of endometriosis-group Ia (without surgical history, n=16), group Ib (with a surgical history, n=27). Results: The mean age was not significantly different between group I and group II. However, AMH and FSH were significantly different between group I and group II ($1.9{pm}1.9$ ng/mL vs. $4.1{pm}2.9$ ng/mL, $p$ <0.01; $13.1{pm}7.2$ mIU/mL vs. $8.6{pm}3.3$ mIU/mL, $p$ <0.01). Furthermore, the number of retrieved oocytes and the number of matured oocytes were significantly lower in group I than in group II. In group II, AMH and FSH as well as age were significant predictors of retrieved oocytes on univariate analysis. Only the serum AMH level was a significant predictor of poor ovarian response in women with endometriosis. Conclusion: Serum AMH may be a better predictor of the ovarian response of COH in patients with endometriosis than basal FSH or age. AMH level can be considered a useful clinical predictor of poor ovarian response in endometriosis patients.