Numerical hydrologic simulation, in combination with synthetic weather generation, was used to assess the sensitivity of the peak flow regime for hourly, daily, and 7-day discharge to hypothetical harvesting scenarios in Redfish Creek in southern British Columbia. Simulation was based on stationary vegetation cover, thus representing potential peak flow regime adjustments during the first few critical years following harvesting. The relative difference between preharvest and postharvest peak discharge quantiles (ΔQT), estimated by comparing GEV distributions fit to the simulated preharvest and postharvest annual maximum series, tended to increase with increasing harvest area; however, the relationship was strongly influenced by the elevation distribution of harvest blocks. Although rainfall was determined to be a factor in roughly three quarters of all peak discharge events, the flood frequency of Redfish Creek is fundamentally a function of the frequency structure of snowmelt runoff. Consequently, changes in runoff synchronization between various elevation bands largely drove the magnitude of ΔQT, and the degree of increased synchronization was found to be greatest following harvesting above H60 (elevation above which lies 60% of basin area) and least following harvesting below H60. As a result, only harvesting schemes that incorporated cut blocks above H60 tended to generate significant (α = 0.05) ΔQT for a wide range of return periods (T), regardless of discharge timescale. Significant ΔQT ranged from 6 to 14% for hourly, 5 to 14% for daily, and 5 to 20% for 7-day discharge in the range of 1.25 ≤ T ≤ 100 years. For a given scenario, hourly and daily ΔQT showed little variation with T, suggesting that small and large discharge events responded similarly to forest harvesting. However, this apparent trend substantially overstates ΔQT for the largest hourly and daily peak discharge events such that the relationship of ΔQT to T is inconclusive for T > 30 years. For 7-day discharge, ΔQT increases with increasing T for 1.25 ≤ T ≤100 years.