Endosseous dental implants are available with various surface characteristics ranging from relatively smooth machined surfaces to more roughened surfaces created by coatings, blasting by various substances, by acid treatments, or by combinations of the treatments. Studies characterizing these implants and surfaces include in vitro experimentation, animal studies, and human clinical trials. Both descriptive and functional testing of the bone‐implant interface includes histomorphometrics and biomechanical testing such as torque removal values and push out/pull out strength. Using these assays to evaluate and compare different surfaces, the data demonstrate that rough implant surfaces have increased bone‐toimplant contact and require greater forces to break the bone‐implant interface compared to more smooth surfaces. The objective of this report was to evaluate publications of human clinical experiences evaluating implant use in patients and to determine if differences existed in success rates of implants with relatively smooth surfaces compared to implants having roughened implant surfaces. Human trials were reviewed to determine the clinical efficacy of implants under various clinical indications. Synopsis tables were constructed and the experiences segregated by implant surface characteristics. Meta‐analyses were performed on all implants in all locations, on implants placed only in the maxilla or the mandible, and, finally, on implants placed in the maxilla compared to implants placed in the mandible. Evaluation of the data revealed that predictably high success rates can be achieved for implants with both rough and smooth titanium surfaces and for hydroxyapatite‐coated implants. When studies were clustered by specific indications or patient populations, rough surfaced implants had significantly higher success rates compared to implants with more smooth surfaces except in the case of single tooth replacements where the success rates were comparable. In general, implants placed in the mandible had significantly higher success rates than implants placed in the maxilla. However, in the partially edentulous patient group, titanium implants with a rough surface had significantly higher success rates in the maxilla compared to the mandible and, in cases of single tooth replacement, success rates were similar in the maxilla and in the mandible as was the case for hydroxyapatite‐coated implants. The documented advantage of implants with a roughened surface in animal and in vitro experiments has been demonstrated in clinical cases when studies were compared in which specific indications or patients were treated. Additionally, implants placed in the mandible have, in general, higher success rates than implants placed in the maxilla, with only a few exceptions noted. These data from human clinical experiences support the documented advantage of implants with a roughened surface in animal and in vitro experimentation and indicate that the magnitude of the advantage is significant for patient care. J Periodontol 1999;70:1523‐1539.