Rice (Oryza sativa L.) cultivated with the direct‐seeded, delayed‐flood production system relies heavily on post‐emergence aerial application of N. The availability of a controlled‐release N fertilizer suitable for preplant application would offer rice growers an alternative N‐fertilization method and reduce the aerial application costs of N fertilization. Our objectives were to determine grain yield and N uptake of rice receiving preplant incorporated polymer‐coated urea (PCU) compared with urea applied preflood at the five‐leaf stage and characterize the N release of two PCUs. Field trials were conducted at five site‐years to evaluate rice performance when fertilized with preplant‐applied Environmentally Smart Nitrogen (ESN) and Duration Type 5 (D5) and preflood‐applied urea across N rates ranging from 0 to 168 kg N ha⁻¹ Nitrogen release from PCU was evaluated in field incubations using a buried‐bag method at two site‐years. Nitrogen release was nonlinear across time and similar between site‐years but different between PCUs. The nonlinear relationships predicted that 75% of PCU N content was released by 36 d for D5 and 25 d for ESN. Nitrogen recovery at panicle differentiation averaged 30% for D5, 26% for ESN, and 72% for urea and at heading averaged 47% for D5, 37% for ESN, and 101% for urea. As the N rate increased, yields increased nonlinearly for rice receiving D5 preplant and urea preflood and linearly for rice receiving ESN preplant. Yield predictions for D5 were always lower than for urea at the same N rates. Results suggest that the N release from D5 and ESN is too rapid for rice cultivated in the direct‐seeded, delayed‐flood method.