Background: Poor response to controlled stimulation is a major problem in ART. Increasing number of mature oocytes and embryos is an important aspect for successful outcome. Objectives: To assess the efficacy and outcome of pregnancy of poor responder women underwent ICSI program using sandwich, conventional antagonist and micro dose protocols. Methods: The randomized controlled trial study included 158 poor responder women underwent ICSI program diagnosed according to POSEIDON classification, were randomly divided into three groups; Microdose protocol N=41, conventional antagonist protocol N=71 and sandwich protocol N=46.The program performed by using high dose gonadotrophins from second day of cycle till administration of HCG. The primary outcomes were number of retrieved oocytes and pregnancy rates. Secondary outcomes were duration of stimulation, total amount of gonadotrophins needed, endometrial thickness on day of HCG trigger, maturity of oocytes, number of transferred embryos, fertilization rates, implantation rates and cancellation rates. Results: There was higher endometrial thickness on day of administration of HCG in microdose group compared to sandwich and conventional antagonist groups (p˂0.05%).While there were higher numbers of retrieved oocytes, mature oocytes, number of transferred embryos with higher fertilization rate, implantation rate and lower cancellation rates in sandwich group compared to microdose and conventional antagonist groups (p˂0.05%). While the pregnancy rate was slightly higher in microdose than sandwich group. While there was no statically significant difference regarding stimulation period, total required amount of gonadotrophins and peak estradiol level on day of HCG trigger (p-value >0.05%. Conclusions in poor responders the sandwich protocol improved the outcome of pregnancy.