Rice (Oryza sativa L.) ratooning has attracted attention due to its higher yield potential compared with conventional methods. Here, field experiments were conducted in southwestern Japan in 2019 and 2020 to determine the effects of cultivar (‘Takanari’, an indica cultivar, and ‘Koshihikari’, a japonica cultivar) and harvest time of the first crop (early and late) on the grain yields of the first and second crops in rice ratooning. For Koshihikari in 2019, grain yield of the second crop did not differ between harvest times. However, Takanari produced a lower grain yield at late harvest. Takanari had fewer spikelets, fewer filled spikelets, and a lower 1,000‐grain weight at late harvest. However, for Takanari in 2020, grain yield did not differ between harvest times. In 2019, the stubble nonstructural carbohydrate (NSC) content did not increase from early to late harvest, whereas the stubble leaf area index (LAI) decreased between harvest times; this difference was more apparent in Takanari than in Koshihikari. Contrastingly, in 2020, the NSC content increased from early to late harvest, but the stubble LAI did not differ between harvest times. Thus, the lower grain yield of late‐harvested Takanari in 2019 might be attributed to the lower stubble LAI resulting in fewer spikelets. In contrast, the maintained grain yield from early to late harvest in 2020 might be attributed to the maintenance of spikelet numbers via the higher stubble NSC content. Therefore, the sufficient stubble LAI and NSC could be essential for improving grain yield of the second crop. Core Ideas Takanari produced a lower grain yield of the second crop at late harvest in 2019. Takanari had fewer spikelets at late harvest due to the lower stubble leaf area index (LAI) in 2019. For Takanari, grain yield of the second crop did not differ between harvest times in 2020. Takanari did not have fewer spikelets at late harvest due to the higher nonstructural carbohydrate (NSC) content in 2020. The sufficient stubble LAI and NSC could be essential for improving grain yield.