This work evaluates the effect of different slow-release fertilizers and nitrification inhibitors (NI) on N-use efficiency, grain yield and N-2 fixation in rice fields of Valencia (Spain) during three consecutive crop seasons (1998-2000). Eight N sources [ammonium sulphate, urea, polymer-coated urea (PCU 32% and 40% N), sulphur coated urea, isobutylidene diurea (IBDU), ammonium sulphate nitrate (ASN) plus dicyandiamide and ASN plus dimethyl pyrazole phosphate, were applied at 120 kg N ha(-1) and at 2 or 15 days before flooding (DBF) during 1998. Another experiment was based on the use of urea blended with PCU (40% N) at four ratios (1:0, 3:1, 1:1, 1:3) and applied at 15 DBF and at four rates (70, 95, 120 and 145 kg N ha-1) during 1999 and at only one rate (120 kg N ha(-1)) during 2000. Both experiments also included a control (no N). The results showed that, when applied shortly before flooding, PCU (32% and 40% N) and IBDU application improved biological N-2 fixation compared to the conventional fertilizer application, with or without NI, reaching similar values to those observed in the absence of N fertilizer. Slow release fertilizers, particularly PCU 40% N applied basally before flooding, were the best N source for grain yield and improved recovery efficiency. Differences among N sources were only significant when the flooding was delayed for 15 days after fertilizer application.