•We used two independent approaches to identify the beginning and end of the critical period for seed number determination in soybean.•Critical period for seed number determination occurs during the R3-R6 phase.•Seed number was related to biomass accumulation during the critical period.•Accumulated biomass is a better predictor of seed number than crop growth rate or duration alone. Seed number (SN) is positively associated with aboveground dry matter (ADM) accumulation in most crop species. In soybean, there are discrepancies among studies relative to the start and end of the critical period for seed number determination. The objective of this study was to more precisely identify those two stages. To do so, we followed two approaches: (i) analyses of experimental data where differences in SN occurred as a result of variation in crop growth due to weather and management practices and (ii) review of published experiments where soybean was subjected to artificial stresses during specific crop phases. Following the first approach, we analyzed the relationships between SN versus ADM accumulation and its components, crop growth rate (CGR) and duration, for different reproductive phases using data from experiments that portrayed a wide range of environments and management practices. In the second approach, we compared SN values in stressed versus non-stressed treatments. Accumulated ADM during the R3 to R6 phase explained ca. 70 % of the variation in SN across experiments. The CGR during R3-R6 was also a good predictor of SN, but it had lower explanatory power compared with accumulated ADM, because CGR did not account for changes in duration of critical period across experiments. Review of published data, wherein soybean crops were subjected to artificial stresses during different reproductive phases, confirmed that the R3-R6 phase is a critical phase for SN determination in soybean. Key findings of this study were: (i) critical period for SN determination in soybean occurs during the R3-R6 phase and (ii) accumulated ADM during the critical period is a better predictor of SN compared with CGR because it accounts for differences in the duration of the critical period as a result of weather and management.