Predictions and quality management issues for environmental protection in river basins rely on water-quality models. These models can be used to simulate conditions in or near the range of the calibrated or verified conditions. In this respect, estimation of parameters, which is still practiced by heuristic approaches (i.e. manually), seems to be the point where the attention needs to be focused. The authors' research group has developed a systematic approach for dynamic simulation and parameter estimation in river water quality models, which has eliminated the cumbersome trial-end-error method. This study reports a user-interactive software named as RSDS (River Stream Dynamics and Simulation) for the implementation of the suggested technique, and provides a comparative investigation of the suggested modelling approach against a well established and worldwide known water quality software, QUAL2E. Experimental data collected in field observations along the Yesilirmak river basin in Turkey were checked against the predictions from both software programs. The results indicated that much better agreement with the experimental data could be obtained from RSDS compared with QUAL2E. Thus, the systematic procedure suggested in the present work provides an effective means for reliable estimation of the model parameters and dynamic simulation for river basins, and therefore, contributes to the efforts toward predicting the extent of the effect of possible pollutant discharges in river basins. (c) 2007 Elsevier Ltd. All rights reserved.