The time it takes water to travel through a catchment, from when it enters as rain and snow to when it leaves as streamflow, may influence stream water quality and catchment sensitivity to environmental change. Most studies that estimate travel times do so for only a few, often rain-dominated, catchments in a region and use relatively short data records (<10 years). A better understanding of how catchment travel times vary across a landscape may help diagnose inter-catchment differences in water quality and response to environmental change. We used comprehensive and long-term observations from the Turkey Lakes Watershed Study in central Ontario to estimate water travel times for 12 snowmelt-dominated headwater catchments, three of which were impacted by forest harvesting. Chloride, a commonly used water tracer, was measured in streams, rain, snowfall and as dry atmospheric deposition over a 31 year period. These data were used with a lumped convolution integral approach to estimate mean water travel times. We explored relationships between travel times and catchment characteristics such as catchment area, slope angle, flowpath length, runoff ratio and wetland coverage, as well as the impact of harvesting. Travel time estimates were then used to compare differences in stream water quality between catchments. Our results show that mean travel times can be variable for small geographic areas and are related to catchment characteristics, in particular flowpath length and wetland cover. In addition, forest harvesting appeared to decrease mean travel times. Estimated mean travel times had complex relationships with water quality patterns. Results suggest that biogeochemical processes, particularly those present in wetlands, may have a greater influence on water quality than catchment travel times.