Along with rapid population growth and increasing human needs, water supply is of great importance. Usage and management of water resources, requires an understanding of hydrologic phenomenon which runoff is considered as one of the most important phases of this cycle. In order to employ appropriate management policies and operational decisions for best practices of water resources, runoff prediction as the reservoir input, is of special significance. Different models with different structures can be used for estimation and prediction of the river flow and provide/present convincing results. In this thesis, HEC-HMS model with SMA loss calculation method which is capable of simulating the inflow continuously (time sequenced), was selected as a river basin model for runoff prediction. This model is utilized to simulate the daily Zayanderud reservoir inflow at Ghale Shahrokh station. Daily discharge data for calibrating the model was obtained for 4 years (from 1992 to 1996) and consequently for validation from 2002 to 2006. HEC-HMS model, in the case where there is no distinction between runoff resulted from rainfall or snow, has failed to provide acceptable results, and the peak flows were estimated incorrectly. The results showed that by considering the snow melt option and incorporated that into the model, the HEC-HMS model results are improved and model estimates of pick flows are much comparable to observed figures. HEC-HMS model results considering different goodness of fit tests are: R 2 = 0.84, Nash-Sutcliffe = 0.77, Bias = 4.3cms, PTVE = 14.9%, RMSE = 12.1cms and PW-RMSE = 14.1cms. Comparing statistical and graphical HEC-HMS (Along with continuous SMA loss method) model with other conceptual and black boxes models that were evaluated in the same basin represents the model capabilities in integrated modeling of runoff in the daily scale. Evaluation of model uncertainty using Monte Carlo analysis and use of the most influential model parameters including the maximum soil infiltration, the impervious percentage and the soil percolation with an uncertainty of 50% was performed. The results indicated that the model is fairly reliable if it is subjected to considerable changes of input parameters. Key words : Zayanderud basin, Rainfall-runoff simulation, SMA loss method, HEC-HMS, Monte Carlo analysis