The Rif mountains in northern Morocco experience frequently geomorphological hazards, including landslides that hamper urban management.This paper aims to develop an objective approach for the assessment of landslide susceptibility in the city of Al Hoceima based on a bivariate probabilistic model (weight of evidence). This approach allows analyzing the relationships between the spatial occurrence of landslides and the different natural and anthropogenic parameters that tend to accentuate and aggravate the formation of this phenomenon. The purpose of our work is landslide hazard zoning while ensuring a better zoning of the landslide susceptibility with a good prediction. The results of the Receiver Operating Characteristic (ROC) curve show that the comparison of the preferred landslide susceptibility map with the inventory map gives a good predictive capability (AUC= 0.902). The obtained map shows that high to very high susceptibility zones contain 69% of the total landslides inventoried. The mapping products of this work can be considered as a major contribution to the elaboration of urban development plans in the city of Al Hoceima