Resorbable membranes of poly(lactic acid) and poly(glycolic acid) (PLA/PGA) were compared to nonresorbable expanded polytetrafluoroethylene (e-PTFE) membranes in the treatment of defects around titanium dental implants placed in postextraction sockets. Two partially edentulous and three completely edentulous patients requiring implant-supported restorations participated. Sixteen Brånemark implants were placed into extraction sockets and covered with modified titanium cover screws, called harvest cover screws, which allow tissue biopsy at second-stage implant surgery. Seven defects were treated with PLA/PGA membranes, five were treated with e-PTFE membranes, and four were left untreated (control sites). After 6 months of healing, the harvest cover screws were retrieved and processed for light microscopy examination together with the regenerated tissues. Very little or no bone formation was detected in control specimens. The e-PTFE membranes were found to be the most effective barrier material, in that denser and a greater amount of regenerated bone was found. The PLA/PGA membranes produced some bone regeneration when compared to control sites, but to a lesser extent compared to e-PTFE sites. Keywords: dental implant, expanded polytetrafluoroethylene membrane, guided tissue regeneration, poly(glycolic acid) membrane, poly(lactic acid) membrane