Diminished ovarian reserve predicts disappointing pregnancy chances. Little data have concerned predictor for in vitro fertilization (IVF) outcome in women with diminished ovarian reserve. To investigate the association of basal testosterone (T) levels with ovarian response and in vitro fertilization (IVF) outcome in women with diminished ovarian reserve. Retrospective cohort study. We reviewed data of patients underwent their first IVF cycle, without polycystic ovary syndrome (PCOS) or endometriosis, from 2009-2010. Diminished ovarian reserve was defined as basal FSH value >10IU/L or FSH:LH ratio >3. The association of basal FSH, LH, E2, T levels with ovarian response and IVF outcome were analyzed with SPSS statistical package. We reviewed 1598 IVF cycles. Complete clinical data of 223 women with diminished ovarian reserve were analyzed. The mean age of patients was 34.51 ± 5.12 years old. Antral follicle count (AFC) was 6.59 ± 3.66. Levels of basal serum FSH, LH, E2 and T were10.55 ± 3.90IU/L, 3.57 ± 1.88 IU/L, 0.15 ± 0.10 nmol/L and 1.37 ± 0.61 nmol/L. The average total dose of gonadotropins was 2666.13 ± 960.33IU. The numbers of large follicles (>14 mm) on HCG day, oocyte retrieved and high-quality embryo were 6.20 ± 2.94, 7.76 ± 3.97 and 1.34 ± 1.37. Endometrial thickness on hCG day was 1.02 ± 0.25cm. The pregnancy rate was 29.60%. Basal T was correlated with the numbers of large follicles on HCG day (r = 0.132, P=0.049), numbers of oocyte retrieved (r = 0.144, P=0.031), two pronuclei (r = 0.137, P=0.041). On the other hand, significantly negative correlations were observed between basal T and basal FSH (r = -0.16,P=0.017), total dose of gonadotropins (r = -0.154, P=0.02). A testosterone level of 1.12nmol/L was shown to predict pregnancy outcome with a sensitivity of 83.3% and specificity of 42.3%, P=0.039. In women with diminished ovarian reserve, basal T was a predictor for ovarian response and IVF outcome.