The brown creeper (Certhia americana) is one of the forest bird species most sensitive to partial harvesting in North America. We examined the detailed response of this species and its food (bark-dwelling invertebrates) during the 3^rd and 4^th year after experimental selection harvesting (30–10% basal area removal) in northern hardwood forest. Relative to control plots, nest densities in treated plots were ca 50% lower. Because the density of nesting substrates was not significantly lower in treated plots than in controls, we investigated whether foraging substrates could be the limiting factor. Specifically, we tested for a treatment effect on 1) the abundance and species composition of bark invertebrate assemblages; 2) the biomass of bark invertebrates per unit area; and 3) the frequency of food provisioning. As predicted, treatment had a significant negative effect on food provisioning rate, though not on invertebrate biomass, when accounting for year effects. There was also no evidence for a treatment effect on the structure of bark invertebrate assemblages, which was mainly influenced by cumulative degree days. Selection harvesting thus appeared to reduce the amount of food delivered to brown creeper nestlings, unless greater amounts of food were delivered per feeding trip in treated plots. The lower density of foraging substrates in treated plots (77 versus 112 stems ha^-1 in controls) may require that adults perform longer foraging trips. Future studies should determine whether this extra effort has short- or long-term consequences for adults and nestlings.