Colorado’s Front Range forested watersheds provide municipal water supplies for downstream communities. Many of these watersheds have been affected by wildfires and subsequent runoff, erosion and sedimentation of waterways. Natural resource managers need information on the frequency and duration of post-fire runoff and erosion, particularly in areas that drain to municipal water supplies. The objectives of this research are to: (1) identify whether or not runoff and erosion can be predicted by rainfall thresholds within three recent Colorado Front Range fires, (2) examine whether thresholds change with time since burn (post-fire years 0-4), spatial extent (plot-, hillslope- and watershed-scale), and post-fire treatments (i.e., mulch), and (3) develop a tool for Colorado to estimate the frequency of threshold exceedance, and therefore runoff and erosion events, in future fire areas. We identified minimum thresholds of 60-minute rainfall intensities (MI60) in the range of 0-22 mm hr-1 for untreated sites for years 0-2 post-fire, using subsets of data separated by fire and spatial scale; when all scales and fires were merged, these thresholds were in the range of 7-8 mm h-1. Storms with rainfall intensities exceeding these thresholds generated surface runoff and erosion. Thresholds predicted 56-100% of post-fire runoff events with an average prediction accuracy of 93%. For hillslopes, threshold rainfall intensities in the first two years post-fire were similar for the High Park, Hayman, and Bobcat Fires; MI60 rainfall in the range of 7-12 mm h-1 predicted >85% of erosion events across all three fires combined. Thresholds for runoff increased substantially during the third year after fire, when MI60 ranged from 8-22 mm hr-1. Although many factors may change the value of rainfall thresholds for runoff, we found that it is difficult to isolate the effects of a single factor. Consequently, we did not detect clear differences in rainfall thresholds with and without post-fire treatments at the plot-scale, but we did detect the effects of treatment at the hillslope-scale, where treatments increased threshold values on average 1 mm hr-1 relative to untreated areas. Effects of time since burn on thresholds were detected only in the Bobcat and High Park Fires. Results indicate that spatial scale can change thresholds for runoff, but the direction of change is not consistent from plot to watershed scale. Most of the rainfall thresholds identified have less than a 1-year return interval, indicating that post-fire runoff and erosion is likely to occur several times per year during the first two years after fire. The frequency of threshold exceedance increased with increasing elevation from 1500-2100 m; decreased from 2100-2300 m, and was relatively consistent with elevation above 2300 m. Frequency analyses indicate that rain storms with 60-minute intensities of 4 mm h-1 occur between 6 to >10 times per summer in Colorado, and events with intensities between 5-7 mm h-1 occur between 2 and 6 times per summer. Understanding the likely frequency of rainfall events that will cause runoff and erosion after fire will help resource managers plan for post-fire flooding or sediment problems and prioritize treatments to those areas with lower thresholds and higher frequencies of threshold exceedance.