The cutting of all lianas prior to logging is a reduced-impact logging technique that is predicted to reduce liana proliferation in logging gaps. This study compares liana abundance and species composition in gaps created during conventional and reduced-impact logging in a forest of the eastern Brazilian Amazon. Logging treatments were conducted in side-by-side plots. Shortly following logging 50-m2 plots were located in the approximate centers of four single treefall and four multiple treefall gaps in each logging area. Six years following logging there were 40% fewer climbing lianas in reduced-impact gaps than in conventional logging gaps. In both logging areas multiple-tree gaps had higher liana densities and a higher proportion of lianas recruiting from seed than single tree gaps where sprouts from cut or fallen lianas were more common. The mean number of liana species encountered per plot did not differ among treatments nor was there any significant difference in species diversity (Fisher\s ) between logging treatments. The results of this study suggest that pre-logging liana cutting can significantly reduce post-logging liana proliferation in gaps with no discernible negative impact on the species diversity of regenerating lianas.