The study was initiated in 1977 on a Typic tropofluvent soil at the Cropping Systems Research Centre, Karamana, on a rice- rice (Oryza sativa cv. Aiswarya, medium duration) cropping system . The experiment was laid out in a partially confounded factorial design with three levels of N (40,80 and 120 kgha -1), 3 levels of P2O5 (0, 40, and 80 kgha -1), and two levels of K2O (0 and 40 kgha -1). op. After 30 years of cropping grain yield and dry matter production were markedly reduced in treatments continuously depleted of P, but with adequate N and K. The percent increase in grain yield over control increased with incremental additions of N and P but the magnitude of increase was more for P levels. The agronomic efficiency was higher at lower N and P levels but nutrient use efficiency decreased with increasing N, P and K levels. Skipping P significantly decreased N, P and K uptake. For each ton of grain produced, the total nutrient uptake increased with an increasing application levels. The N balance was always negative but phosphorus balance was negative only in control plots and increased to 13.52 kgha-1 annum-1 at P80. K balance was negative for both control and K added plots. Thus imbalanced and inadequate fertilization affects the yield response of wetland rice to nutrients. Though P is not generally a limiting nutrient in the acid entisols of India, its long term depletion was found to affect the crop response to N and K as well.