The report presents the renewal process model—the model based on the random process theory. The model is capable of predicting the timing of a landslide activation as well as of obtaining dynamic parameters for that. The model operates with a geologically uniform area, where the landslide activation is determined by local factors such as storm rainfall and depergelation (melting permafrost). The approach allows minimizing the duration of a monitoring period sufficient for a landslide forecast. We used the landslide database for the area of Seattle to verify the model. Particularly, the model employed the data on shallow landslide occurrences. The model verification showed the distribution of the time between two successive activations has an exponential behavior at significant level of 0.95, while the time distribution between the latest date in the database and the latest time of activation has a lognormal behavior at significant level of 0.95.