– Study of weather prediction is a challenge that is always interesting to study. Although there was several methods of weather prediction, but the results have not provided good accuracy. The use of fuzzy logic has been proven by scientists to be applied to the expression of uncertainty, it is not clear and qualitative from a system. This method produces lower error percentage. This study aims to develop opportunities rainfall prediction algorithm using fuzzy logic type Mamdani and Sugeno, and compare the model results to determine the accuracy of the prediction results. The data are input parameters in the scale meteorological processes that influence the occurrence of rain. Parameters are classified into several categories to make it easier to make the rules (IF-THEN rules). The model was built and the results were tested in the wet season, dry season and transition season with observational data. Results of verification states that Mamdani type is more reliable as a weather prediction models with the accuracy percentage of 77%, 80% and 84%. While the percentage of accuracy obtained Sugeno type 32%, 63% and 42%. Keyword : Fuzzy Logic, Weather Prediction, Mamdani, Sugeno, Meteorology Scale