The curve number (CN) method is one of the most common methods for estimating runoff in watersheds. The CN value depends on the soil infiltration, vegetation cover, and antecedent soil moisture content. According to the antecedent soil moisture content, three types of CN can be used: CN for dry soil moisture condition, CN for medium soil moisture condition, and CN for wet soil moisture condition. However, the determination of CN from the vegetation cover and soil infiltration needs high accuracy. In this study, 63 measured rainfall-runoff events in the Baghan watershed and 34 measured rainfall-runoff events in the Booshigan watershed were used for estimating runoff with seven conditions: 1) Calculating the CN value based on soil information and vegetation cover in different parts of the watershed. 2) Calibrating the surface storage coefficient of the runoff estimation equation by considering the medium condition for all events. 3) Calibrating the surface storage coefficient of the runoff estimation equation by considering the real soil moisture condition (dry or medium or wet) of each event. 4) Considering the average of observed CN in all measured rainfall-runoff events. 5 to 7) Calculating the relationship between the observed CN and measured rainfall as linear, power, and standard equation. For mentioned conditions, 48 measured rainfall-runoff events in the Baghan watershed and 26 measured rainfall-runoff events in the Booshigan watershed were used for calibrating the results, and the remained measured data in each watershed were used to evaluate the results. The results in two watersheds demonstrated that the linear, power, and standard conditions were better for runoff estimation. In the three mentioned conditions, the CN value depends on the measured rainfall data instead of using the soil infiltration information and vegetation cover of the watershed. Also, the results showed that it is not suitable to use the common method for determining the CN value (based on soil infiltration and vegetation cover of the watershed), and then estimating the runoff amount.