Weather variability poses threats to rural crop producers in Tanzania. This research aimed to find the impact of weather variation on the growth stage and yield of rice in Tanzania. The analyses were done using rice yield data and weather variables from 1981-2017. The approaches used were; decomposing rice yields into yield tendency and yield weather, stepwise integral regression for identification of significant yield model, and applied Fisher’s meteorological regression and Chebyshev polynomial function to compute coefficients for weather factors. From the results, other than the non-natural factors, rainfall, maximum and minimum temperature, and sunshine significantly affect rice yield from sowing to harvest stage. The effect of rainfall, sunshine, maximum and minimum temperature coefficients on the rice yield differ by growth stage. An increase of 1 millimeter of rainfall at the sowing-seedling stage increased rice yield by 2.7 kg/ha. In the sowing-seedling stage, the temperature had a stronger positive influence on the rice yield as with every 1°C in average maximum temperature increased the rice yield by 674.1 kg/ha. The minimum temperature coefficient had stronger influences in the vegetative, tillering-booting stages, thus, with 1°C increase in average minimum temperature, the rice yield increased by 70.1 kg/ha and 420.7 kg/ ha respectively. In the flowering-grain formation stage, the maximum temperature had a greater influence on rice yield, that is, as 1°C increased, the rice yielded increased by 674.7 kg/ha. The sunshine duration had a higher influence on the harvesting stage. Increased 1-hour duration of sunshine increased rice yield by 495.95 kg/ha. Finally, a meteorological rice model, which could be used for rice yield forecasting in the region, was developed.