Abstract Study question Is follicular phase or luteal phase stimulation more effective in progestin primed double stimulation protocol in poor responders? Summary answer The luteal phase stimulation was more effective than follicular phase stimulation in the progestin primed dual stimulation protocol in poor responders. What is known already Poor ovarian responders involve 9–24% of patients undergoing in vitro fertilization (IVF), with the management of these patients posing an everyday practical challenge. Owing to the follicular waves theory, two or three waves of folliculogenesis were detected during the interovulatory interval, two consecutive ovarian stimulation protocols could be initiated in the follicular phase and the subsequent luteal phase. This is an effective way to increase the number of oocytes and embryos obtained over a relative short period of time. Furthermore, there is increasing evidence that progestins can be a reliable method of preventing premature LH surge during ovarian stimulation. Study design, size, duration This was a randomized controlled study performed at the Infertility and Assisted Reproduction Unit of ElShatby Maternity Hospital, Alexandria University and Madinah Fertility Centre. The study was conducted on 90 infertile women with poor ovarian response as defined by the Bologna criteria. 7 patients were excluded due to protocol violation, loss to follow-up and patient withdrawal. The study commenced in September 2020 and was finalized in September 2021. Participants/materials, setting, methods Group 1 45 patients were given the progestin primed double stimulation protocol. Dydrogesterone was started from the first day of the ovulation induction till the end of luteal phase stimulation. Embryo transfer was scheduled in a subsequent cycle. Group 2 45 patients were given the flexible GnRh antagonist protocol in 2 cycles. The best blastocyst from either cycle was transferred in the second cycle. Main results and the role of chance When comparing the follicular and luteal phases, the M2 oocytes number retrieved in the luteal phase was statistically significantly higher than the follicular phase, with the median number of 4 oocytes against a median number of 2 oocytes (P value 0.001). The fertilization rate was significantly higher in the M2 oocytes of the luteal phase stimulation (P value 0.04). The resultant embryos from the luteal stimulation phase were statistically significantly higher than the follicular phase, with the median number of 3 embryos against a median number of 1 embryo (P value <0.001). However, we found that the days of controlled ovarian hyperstimulation were statistically significant higher in stimulation of the luteal phase with the median of 12 days against the 10 days seen in follicular phase stimulation (P value 0.002). Consequently, the total dosage of gonadotropins was statistically significant higher in the luteal phase with a P value of 0.007. After Subgroup analysis, Poseidon Group 4 patients demonstrated that a significantly higher number of oocytes were retrieved after the progestin primed double stimulation than after two waves of flexible GnRh antagonist stimulation (median of 12 oocytes against a median of 10 oocytes with P value 0.01). Limitations, reasons for caution Firstly, preimplantation genetic testing was not performed to detect euploid embryos which could be vital when comparing both the phases and the protocols. Secondly, despite the importance of the live birth rate as an outcome measure, we did not enroll this measure. Wider implications of the findings In the light of our findings, we recommend the usage of progestin primed dual stimulation protocol in poor responder patients. Additionally, we call for multicentric randomized controlled research to verify this protocol as a first-line treatment protocol in poor responder patients. Trial registration number ClinicalTrials.gov ID: NCT04537078