The synthetic unit hydrograph method is a popular method for analyzing watershed flood discharge for rivers that do not have observational flood hydrographs. To create flood hydrographs for rivers with no or very few observed flood hydrographs, it is necessary to require data on characteristics or parameters of watershed areas (DAS). The Time to peak Model in this study will consider several parameters including the area of the watershed (A), the length of the main river (L), the length of the river from the center of the watershed to the outlet (Lc), river slope (S), watershed roughness (n), factor the shape of the watershed (Fb), as well as the fractal characteristics of the watershed in the form of river branching ratio (RB) and river length ratio (RL). To get the accuracy of the time to peak model and to get a flood hydrograph that can represent the prototype, it is necessary to do some statistical analysis. The time to peak model predicted using linear regression analysis produced the time to peak equation (Tp) as a function of watershed area (A) and river length from the center of the watershed to outlet (Lc). This equation has a good level of accuracy with a correlation coefficient of 0.893; The coefficient of determination is around 0.797 and Adjusted R2 is around 0.746. The model has also met the requirements of the classical assumptions including linearity test, residual normal test, heteroscedasticity test, autocorrelation test (indicated by the Durbin Watson value of 1.895 with the interpretation of no autocorrelation) and multicollinearity test. Model validation has NSE value of 0.731 (good), RMSE value of 0.482 and MAE value of 0.390 (both RMSE and MAE values are close to zero). Model verification has NSE value of 0.758 (very good), RMSE value of 0.394 and an MAE value of 0.320 (both RMSE and MAE values are close to zero).