Rainfall-induced landslides occur more often with climate change, resulting in destruction and loss of lives in many parts of the world. Predicting such landslides is paramount to maintain local infrastructure and the well-being of communities. Several models have been proposed in the past years; however, most of them only work well when being applied to local site conditions for which they were developed. This work describes a new approach to estimate a model parameter related to shear strength conditions of local soils. To validate the theoretical concepts of a model, a series of shear box and undrained triaxial tests were conducted on soil specimens prepared at different values of water content. The refined model was then applied to three landslide sites to estimate the stability of slopes against past rainfall events. The obtained results showed more accurate predictions of landslide occurrence compared to the existing models. This paper presents and discusses field and laboratory data as well as the outcomes of numerical analysis.