Riparian forests have cool and humid microclimates, and one aim of leaving forested buffer strips between clear-cut areas and streams is to conserve these microclimatic conditions. We used an experimental study set up of 35 streamside sites to study the impacts of buffer strip width (15 or 30 m) and selective logging within the buffer strips on summer-time air temperature, relative air humidity and canopy openness 12 years after logging. The buffer strip treatments were compared to unlogged control sites. We found that 15-meter buffer strips with or without selective logging and 30-meter buffer strips with selective logging were insufficient in maintaining temperature, relative humidity and canopy openness at similar levels than they were in control sites. In contrast, 30-meter buffer strips differed only little from control sites, although they did have significantly lower mean air humidity. Microclimatic changes were increased by southern or southwestern aspect of the clear-cut, and by logging on the opposite side of the stream. We also tested how the cover of three indicator mosses (Hylocomium splendens, Pseudobryum cinclidioides and Polytrichum commune) had changed (from pre-logging to 12 years post-logging) in relation to post-logging air temperature, relative air humidity and canopy openness. We found that each of the species responded to at least one of these physical conditions. Air humidity was the most significant variable for explaining changes in the cover of the indicator moss species, suggesting that the changes in this microclimatic component has biological impacts. We conclude that to preserve riparian microclimatic conditions and species dependent on those, buffer strips should exceed 30 m in width, and not be selectively logged. Wider buffer strips are required if the clear-cut is towards south or southwest, or if the two sides of the stream are logged at the same time or during subsequent years.