Studies have suggested that the need for site-specific fertilizer application to reduce waste of resources as soil nutrients varies across the Malaysian rice fields. The present study was aiming to determine the physiological and yield responses of five rice varieties (MR269, MR297, MR220CL2, MR219, and UPUTRA) treated with three nitrogen (N) rates under farmer’s field condition. The experiment design was a split-plot randomized complete block design, in which N rates, i.e. low (76 kg N ha-1), farmers’ practice (109 kg N ha-1) and high (142 kg N ha-1), were the main plot while rice varieties were the sub-plot. In general, harvestable panicle yield including the grain (at 14% moisture content) of all varieties was between 5.4 and 7.4 t ha-1 under all N treatments. Among all varieties, MR220CL2 recorded significantly higher yield, irrespective of N treatments. The physiological responses of rice varieties to N treatments, however, were mostly non-significant except for panicle biomass. Specifically, each variety recorded different biomass partitioning percentages for different organs at different growth phases. In the case of N treatments, there was no significant difference in the yield between 30% low and 30% high N rates as compared to farmer’s practice at harvest. Hence, it could be suggested that farmers may reduce N application of about 30% without significant reduction in harvestable panicle yield for this specific plot. However, the potential N uptake might also be affected by unaccounted factors such as the availability of micronutrients and planting density.