Increased grain yield in response to high rates of application of nitrogen (N) fertiliser is often limited by increased spikelet sterility, particularly under low temperature conditions in the New South Wales (NSW) rice industry. In 3 field experiments, different N rates were applied for different sowing dates to investigate the interaction between N rate and temperature during microspore development on spikelet sterility and grain yield. In one experiment the effect of water depth on spikelet sterility was also investigated. Engorged pollen production, spikelet sterility, and yield and its components were recorded. Application of N affected a few different processes that lead into spikelet sterility. Application of N at both pre-flood (PF) and panicle initiation (PI) significantly reduced the number of engorged pollen grains per anther, which was negatively correlated with spikelet sterility. Application of N and low temperature during microspore development with the absence of deep water also decreased pollen engorgement efficiency (the percentage of pollen grains that were engorged). Application of N further increased spikelet density, which, in turn, increased both spikelet sterility and grain yield. The combined effect of spikelet density and low temperature during microspore development explained the 44% of variation in the number of engorged pollen grains per anther. Grain yield was decreased by low temperature during microspore development in the shallow water when N was applied. Spikelet sterility as a result of late sowing was strongly correlated with minimum temperature during flowering. It is concluded that N application reduced pollen number per anther as a result of increased spikelet density, and this made the spikelets more susceptible to low temperature, causing increased spikelet sterility.