A recent series of bark beetle outbreaks in the Rocky Mountain region of the U.S. is the largest and most intense ever recorded. Factors contributing to tree mortality from bark beetles are complex, but include aspects of forest stand condition. Because stand conditions respond to forest management, evaluating bark beetle-caused tree mortality and changes in forest structural attributes in areas previously subjected to management not only improves mechanistic understanding of beetle-caused changes in forests, but also improves prediction of future bark beetle responses to management regimes. We retrospectively assessed mortality of lodgepole pine (Pinus contorta), Engelmann spruce (Picea engelmannii), and subalpine fir (Abies lasiocarpa), and stand structure in two watersheds in south-central Wyoming, U.S.A. following outbreaks of mountain pine beetle (Dendroctonus ponderosae), spruce beetle (Dendroctonus rufipennis), and western balsam beetle (Dryocoetes confusus). One watershed received 240 patch cuts (mean area 1.5ha), a type of group selection cut, six years before the beetle outbreaks began; the other watershed received no active management (control). We conducted surveys of forest vegetation attributes over 27yrs, during pre-harvest, post-harvest, and post-outbreak periods. After the outbreak, lodgepole pine and Engelmann spruce mortality increased with increasing bole diameters and basal area of each species, but patterns of mortality were influenced by patch-cutting. Large-diameter trees in or near patch cuts tended to escape attack by bark beetles. Away from patch cuts (⩾15m), mortality of smaller lodgepole pine was higher compared to the control watershed. Based on our observed patterns of tree mortality, we hypothesize a changing pattern of host selection (i.e., selection for smaller trees) was influenced by stand conditions that created more suitable conditions for bark beetles in areas between patch cuts in the treated watershed. Snag density increased from pre-harvest to post-outbreak periods, but log density was similar, suggesting most dead trees remained standing at the time of data collection. Canopy cover did not decrease as expected, and ground cover did not change substantially from pre-harvest to post-outbreak periods. Patch-cutting improved survival probability of large-diameter lodgepole pine and Engelmann spruce during outbreaks of multiple species of bark beetle, although reduced losses were only realized for trees in or near (⩽15m) patch cuts. However, during intense, broad-scale tree mortality events, these benefits may be important in reducing the loss of mature trees to bark beetles and promoting retention of a larger cohort of mature trees post-outbreak.