In this study, our objective was to determine the effect of adding estradiol hemihydrate (E2) to progestin (P) for luteal phase support on pregnancy outcome in in vitro fertilization (IVF) cycles with poor response to gonadotropins. Ninety-five women with poor ovarian response who underwent controlled ovarian hyperstimulation (COH) with gonadotropin releasing hormone (GnRH) agonist or GnRH antagonist plus gonadotropin protocol for IVF were prospectively randomized into three groups of luteal phase support after oocyte retrieval. Group 1 (n = 33) received only intravaginal progesterone gel (Crinone 8% gel). Group 2 (n = 27) and Group 3 (n = 35) received intravaginal progesterone plus oral 2 and 6 mg estradiol hemihydrate, respectively. Main outcome measures were overall and clinical pregnancy rates (PRs) per patient. Serum LH, E2 and P levels at 7th and 14th days of luteal phase were also measured. Overall and clinical PRs were significantly higher in 2 mg E2 + P than P-only group (44% versus 18% and 37% versus 12.1%, respectively). There were no statistically significant differences between 6 mg E2 + P versus P-only and 2 mg E2 + P versus 6 mg E2 + P groups regarding PRs. Addition of 2 mg/day E2 in addition to P for luteal support significantly increase overall and clinical PRs in cycles with poor response to gonadotropins after IVF.