Forest harvesting has been shown to significantly affect freshwater communities in various aquatic ecosystems. Buffer zones are argued to protect the aquatic-terrestrial ecotone to ensure that aquatic ecosystems are not negatively impacted during forest harvesting operations. However, the effectiveness of buffer zones in mitigating potential harmful effects to benthic macroinvertebrate communities in boreal lakes is not well understood. In this study, benthic macroinvertebrate communities were sampled from 21 harvested and reference lakes in the Esker Lakes Research Area, a boreal forest ecosystem in Ontario, Canada. Various indices were used to characterize these communities including: total abundance, Simpson's Index of Diversity, total abundance of different Functional Feeding Groups and total abundance of the most common taxa. The main objective of this study was to compare benthic communities from treatment and reference lakes to determine if forest harvesting impacted benthic macroinvertebrate communities, despite the use of buffer zones. Benthic communities did not consistently or significantly differ between reference and treatment lakes. Two-way Analysis of Variance comparisons indicated that benthic macroinvertebrate communities from clear and intermediate lakes differed substantially between years, regardless of treatment. Some indices of macroinvertebrate communities in boggy lakes suggested that communities were changed specifically in treatment lakes; however; the types of changes (mostly reductions) were not those predicted by existing literature and results from other endpoints, such as water chemistry, did not support an effect due to harvesting. Non-Metric Multidimensional Scaling ordination of communities further supported the conclusion that benthic communities were substantially different between pre-harvest and post-harvest years, but that these differences were not be due to harvesting. The results of this study indicate that a one-year pre-harvest study of benthic communities was not sufficient to properly capture the natural variability of this community and establish a suitable reference condition. The conclusions made from this study are constrained by a number of limitations including industry imposed restrictions on the experimental design, possible inter-year observer bias and significant changes in the weather patterns between years. Despite these limitations, it is clear that the minimal percent catchment harvested and the establishment of a buffer zone ensured that benthic macroinvertebrate communities were not negatively influenced by up to 42% harvesting within the lake catchments at the ELRA.