The experiment was conducted to measure crop yields, nutrient concentration, nutrient uptake and balance by using different nutrient management practices for mustard-mungbean-T. aman rice cropping system in calcareous soil of Madaripur, Bangladesh. Different nutrient management practices were absolute nutrient control (T 1 ); farmer's practice (T 2 ); AEZ based nutrient application (T 3 ) and soil test based nutrient application (T 4 ). The practices were compared in a randomized completely block design with three replications over two consecutive years. The average yield through application of soil test based nutrient (T 4 ) was showed effective to get highest yields of mustard (1530 kg ha ). The same practices (T 4 ) exhibited the greatest nutrients uptake by the test crops. The apparent balance of N and K was negative; however it was less negative and less deficiency detect in T 4 treatment. Positive balance of P observed in all practices except in T 1 . There was a positive S balance (7.60 kg ha -1 ) in T 4 but negative in T 1 , T 2 and T 3 . Zinc balance was found positive in T 3 and T 4 and negative in T 1 and T 2 . Boron balance in the system was neutral or slightly positive in T 1 and negative in T 2 but positive in T 3 and T 4 . Organic matter, N, P, S, Zn and B status in soil was improved by T 4 treatment. The results suggested that the soil test based nutrient application is viable and sustainable for mustard-mungbean-T. aman rice cropping system in calcareous soils of Bangladesh.