In caving mines, the risk of wet muck inrush is critical when water and fine material are present. Wet muck inrushes can result in mine closure as they can be responsible for loss of life, equipment, mining reserves, and productive areas. In this study, a multivariable model based on logistic regression was developed to assess the risk of wet muck entry in drawpoints of block caving mines. Based on data from the Esmeralda mine at El Teniente, the model assesses the risk of wet muck entry for long-term mining plans in an effort to mitigate or delay wet muck entry. The main variables used in this model are the monthly water flow rate, the number of neighboring drawpoints in a wet muck state, the extraction height, the dynamic risk zone, and fine ore entry from older neighboring sectors. The dynamic risk zone and the fine material from older sectors are new variables that had not been considered before. A proposed algorithm models the evolution of the topography to define the risk zones. A gravity flow model is used to simulate the fine material coming from old mine sectors. Results show the multivariable model identifies the occurrence or not of wet muck with 80% accuracy, thus demonstrating it can support long-term mine planning in Block/Panel Caving mines to decrease the hazards associated with wet muck entry.