Although efficient use of N remains a critical constraint to productivity in irrigated lowland rice, a comprehensive database does not exist for the efficiency of on-farm management of N and other nutrients. In 1994, IRRI initiated its Mega Project on Reversing Trends of Declining Productivity in Intensive Irrigated Rice Systems in selected rice production domains of five tropical Asian nations to improve on-farm fertilizer use efficiency and to monitor long-term productivity trends as related to fertilizers and other inputs. Data are reported here for the first crop cycle, the 1994–95 dry season. The indigenous soil N supply (INS) was estimated by aboveground crop N uptake and grain yield (GY) in plots without applied N established in farmers’ fields under otherwise favorable growth conditions. The fertilizer N rate each farmer applied to his/her field surrounding these plots was recorded; GY was also measured in that area. In each domain, GY in unfertilized plots varied considerably among farms, as the range between maximum and minimum values within each domain was at least 2.8 t ha-1, thus of comparable magnitude to mean GY for these plots. Fertilizer N rates varied from 36 to 246 kg ha-1 across all domains, but their lack of relationship to INS contributed to relatively low fertilizer N efficiency and high variability in efficiency among farms. Mean agronomic efficiency (ΔGY/applied N rate) for each domain was only 6–15 kg grain kg-1 N, while values for individual farmers ranged from 0 to 59 kg grain kg-1 N. Initial data on P and K fertilizer management also suggest highly variable applications at suboptimal efficiency. These results indicate the potential for greater fertilizer efficiency from improved congruence between the indigenous soil supply and applied fertilizer and emphasize the need for field-specific nutrient management. Although agronomic efficiency and partial factor productivity (GY/applied N rate) can each be used to describe the efficiency of fertilizer applications, a complete analysis of nutrient management should include both terms, grain yield, fertilizer rates, and native soil fertility.