This work consists of the calibration and validation of the SMAP (Lopes et al., 1982) and Tank Model (Sugawara, 1979) models for simulating monthly flows in the Pianco river basin, which is located in the state of Paraiba. To calibrate and validate both models, which are classified as conceptual, concentrated and deterministic, rainfall, potential evaporation and average monthly flow data from the period 1964 to 1988 were used. The model parameters were automatically calibrated using genetic algorithms. The calibration results considering the Nash-Sutcliffe coefficient (NASH) were 0.73 and 0.60 for the SMAP and Tank Model, respectively. The NASH coefficient was 0.69 for the validation data set using SMAP and 0.45 with Tank Model. Given the effectiveness and superiority of the SMAP over the Tank Model for simulating monthly flows in the Pianco River, it is hoped that this study will provide support for better water management in semi-arid regions.