Two large-scale forestry experiments, in Quebec (Sylviculture et aménagement forestiers écosystémique (SAFE)) and Alberta (Ecosystem Management by Emulating Natural Disturbance (EMEND)), were established in the late 1990s to test the effects of alternative silvicultural strategies (e.g., partial cutting) on biodiversity in northern boreal forests. We collected spiders in pitfall traps 2 years after the application of partial-cutting treatments in deciduous stands at EMEND and 6 years after similar treatments in deciduous stands at SAFE. Although we are aware of the challenges imposed by disparate locations and whole-scale experimental methods, our objective was to compare the effects of partial cutting on spider assemblages (diversity and community composition), and in doing so, to formulate a few general statements. Overall, 98 species (6107 individuals) were collected from Alberta and 86 species (3414 individuals) from Quebec. Of these, 44 species were common to both regions. Ordination and indicator-species analyses revealed a distinct effect of geographic separation: the spider assemblages in deciduous stands within the boreal plains ecoregion of Alberta and the boreal shield in Quebec were distinct. However, the effects of partial cutting on spider assemblages within each project were similar: removal of 25%–33% of trees shifted a characteristic old-growth fauna toward one more typical of clearcuts. Indicator-species analysis also revealed the dominance of wolf spider (Lycosidae) species in clearcuts within both experiments and we present evidence that clear-cutting homogenizes spider assemblages. Old-growth forests contain spider faunas that are easily disrupted by moderate partial cutting. In the face of intense harvesting practices, managing for the maintenance of biodiversity and conservation of spider faunas in northern forests will require retention of old-growth forests.