Phosphorus deficiency in soil and erratic rainfall limit rice production in Northeast Thailand. The objective of this study was to investigate rice’s response to different rates and times of phosphorus fertilizer. A field trial was conducted in a farmer’s field with a split plot design. Main plots were established using three rates: 15, 30, and 60 kg P ha-1. Five types of sub-plot were established as follows: (1) P fertilizer was applied by banding below the seeds’ depth before seeding (BBS), (2) broadcasting and incorporation into the soil before seeding, and (3) topdressing at three different DAS (days after seeding). Fertilizer was spread onto the soil surface at 10, 20, and 30 DAS. Significant grain-yield increases were affected by time or rate of fertilizer application. Grain yields were maximized by a rate of 60 kg P ha-1. Fertilizer application by BBS produced higher grain yield than did other treatments, giving higher root-length density and shoot growth, phosphorus uptake, and greater phosphorus-use efficiency than did fertilizer applied by broadcasting or delayed topdressing. BBS resulted in the highest phosphorus enrichment in soil volume with a rate of 60 kg P ha-1. The high solubility of fertilizer in a localized placement may satisfy the seedlings’ high initial demand of phosphorus and provide sufficient phosphorus availability to allow high productivity. Based on these data, in regions with low soil phosphorus availability where rainfall is also erratic, the management of sub-seed banding placement at the start of a crop at the rate of 60 kg P ha-1 is recommended as the optimum phosphorus fertilizer practice.