Wind energy assessment plays a critical role in the exploitation of wind turbines in many countries worldwide. To evaluate the wind energy potential, probability density functions (pdf) are usually used to describe wind speed distributions. The selection of the appropriate pdf reduces the wind power estimation error. The most widely used pdf for wind energy applications is the Weibull probability density function. In this study, five distributions (Weibull, Gamma, Normal, Lognormal, and Rayleigh distribution) were used to model the wind speed distribution. Used data were recorded at eight stations located on the northwestern coast of Senegal. A detailed comparison of the five approaches is carried out in the present work. The parameters of the distributions were calculated using the maximum likelihood method. The suitability of the theoretical distributions to real distribution was evaluated based on the performance criteria (coefficient of determination R2, root mean square error, and the maximum error between the real and the theoretical of the repartition function). Results indicate that the used distributions show a good fit compared to the real distribution with a coefficient of determination between 88% and 99%. The Weibull distribution performs best for Sine Moussa Abdou, Botla, and Sakhor with a coefficient of determination of 99% for the three sites.