OBJECTIVE: How do numbers of oocytes retrieved per In vitro fertilisation (IVF) cycle impact on the live birth rate (LBR) and multiple gestation pregnancy (MGP) rates? DESIGN: Retrospective observational longitudinal study. SETTING: UK IVF clinics. POPULATION: Non-donor IVF patients. MAIN OUTCOME MEASURES: LBR per IVF cycle and MGP levels against number of oocytes retrieved into subgroups: 0, 1-5, 6-15, 16-25, 26-49 oocytes and 50+ oocytes. Relative risk (RR) and 95% CIs were calculated for each group against the intermediate responder with '6-15 oocytes collected'. RESULTS: From 172 341 attempted fresh oocyte retrieval cycles, the oocyte retrieved was: 0 in 10 148 (5.9%) cycles from 9439 patients; 1-5 oocytes in 42 574 cycles (24.7%); 6-15 oocytes in 91 797 cycles (53.3%); 16-25 oocytes in 23 794 cycles (13.8%); 26-49 oocytes in 3970 cycles (2.3%); ≥50 oocytes in 58 cycles (0.033%). The LBRs for the 1-5, 6-15, 16-25 and 26-49 subgroups of oocytes retrieved were 17.2%, 32.4%, 35.3% and 18.7%, respectively. The RR (95% CI) of live birth in comparison to the intermediate group (6-15) for 1-5, 16-25 and 26-49 groups was 0.53 (0.52 to 0.54), 1.09 (1.07 to 1.11) and 0.58 (0.54 to 0.62), respectively. The corresponding MGP rates and RR were 9.2%, 11.0%, 11.4% and 11.3%, respectively and 0.83 (0.77 to 0.90), 1.04 (0.97 to 1.11) and 1.03 (0.84 to 1.26), respectively. CONCLUSION: There was only limited benefit in LBR beyond the 6-15 oocyte group going to the 16-25 oocytes group, after which there was significant decline in LBR. The MGP risk was lower in 1-5 group.