Water erosion has been a major environmental problem associated to forest harvesting operations in humid and temperate climates. The incorporation of this factor in the decision making process has been complex due to the high variability observed in rainfall intensity and magnitude. This paper presents a methodology for incorporating the temporal variability of rainfall in the planning of forest operations. We used a monthly index based on the R-RUSLE erosivity factor. This index allows knowing the most likely value of soil loss and associating a confidence level. Thus, the forest planner has an estimate of erosion rates related to the operations, and an index that reflects the rainfall variability. With this methodology it is possible to identify, for a given level of soil loss, those months in which there is a higher uncertainty. This methodology was applied to a pine plantation in Santa Barbara, Biobío Region, Chile, where the rainfall erosivity ranges from 16 MJ mm ha^-1 h^-1 to 271 MJ mm ha^-1 h^-1 depending on the month of the year.