Ecological effects of genetically modified plants cannot always be predicted based on knowledge of the plant species or transgene. We studied the effects of transgenic aspen (Populus tremuloides Michaux) with reduced lignin and altered growth phenotypes on the feeding performance of gypsy moth larvae (Lymantria dispar L.) and forest tent caterpillars (Malacosoma disstria Hübner). Developmental trials were conducted using one control line and four separate transgenic lines of aspen. Gypsy moth larvae showed a significant reduction in survival on one high-lignin reduction transgenic tree line relative to all other lines, but weights of surviving larvae were similar across tree lines. Forest tent caterpillars showed similar survival and weights on all tree lines. Trials were also conducted to evaluate whether gypsy moth larvae preferred feeding on high-lignin reduction transgenic aspen lines or control trees. While gypsy moth larvae showed no significant preference between the control line and the transgenic line that caused significant reductions in larval survival during developmental trials, they did strongly prefer transgenic leaves causing no such reductions in larval survival. Because effects on feeding larvae varied among tree lines, we concluded that any potential phytochemical alterations in the transgenic lines could not be directly linked to lignin reduction. Because only one transgenic tree line had a negative effect on the herbivores, we propose that this may be an indirect consequence of transgenic manipulation resulting from the insertion point of the antisense Pt4CL gene in the genome, rather than 4CL suppression or lignin reduction.