To explore the effects of cultivation methods on dry matter production and grain yield of rice under different ecological conditions, we measured dry matter accumulation, distribution, translocation, grain yield and its components in a field experiment with different seedling ages and transplanting methods in Renshou, Pixian and Ya’an using a randomized block experimental design. The results showed as follows. (1) The dry matter accumulation of rice such as dry matter weight per stem or in a population was clearly influenced by ecological condition, seedling age, transplanting method, and there were significant interactions of ecological condition and seedling age on dry matter weight of population from tillering to heading stages. Moreover, the interactions of ecological condition and transplanting method, seedling age and transplanting method, and all of the three factors mainly affected dry matter weight per stem and in a population after heading stage. (2) The diversification of dry matter production characteristics was due to the ecological condition changes. The grain yield was primarily come from the accumulation of photosynthate after heading stage in Renshou, but not significantly correlated with the dry matter translocation from stem-sheath. In Pixian, the dry matter translocation from stem-sheath affected grain yield intensely, rice yield was positively correlated with dry matter translocation ratio from stem-sheath at booting stage (r = 0.775*), but negatively with that at maturity (r = –0.757*), significantly. In addition, the dry matter redistribution from stem-sheath after heading was positively correlated with rice yield in Ya’an. (3) The yield in Renshou increased by 5.52% and 17.65% respectively, compared with that in Pixian and Ya’an. Seedling age and transplanting method had an effect on the yield, that was significantly different under the different cultivation methods. Cultivation methods influenced the yield through changing effective panicles per unit area, seed-setting rate, and 1000-grain weight in Renshou, while by changing effective panicles per unit area and spikelets per panicle in Pixian, and by changing total spikelets and 1000-grain weight in Ya’an. (4) It was suggested that, with the effective regulation of cultivation methods on rice dry matter production, a higher yield would be reached. Overall, the 50 d single seedling with optimized-broadcasting treatment could coordinate the dry matter accumulation at different growth stages to promote the grain yield in Renshou and Pixian. Furthermore, the 50 d single seedlings with hand-transplanting treatment enhanced the dry matter accumulation before heading and the dry matter translocation from stem-sheath after heading, resulting in the extraordinarily increased rice yield inYa’an.