At the neck area of dental implant surface, machined surface (Ms) has been employed to avoid surface contamination. Recently, implants which have roughened surface texture (Rs) at their neck are also available. However, from the viewpoint of soft tissue integration, it remains to be elucidated whether or not surface topography affects the soft tissue attachment around implants. The aim of the present study was to clarify the influence of surface topography on peri-implant soft tissue integration. First, surface roughness of both surfaces was measured. Second, protein adsorption capability on both surfaces was examined. Then, as the rat implant model, titanium implants with each surface were inserted into the maxillae. Horseradish peroxidase (HRP) tracer was applied 4 weeks post implantation to the gingival sulci of implants or natural teeth (NT) to investigate the sealing capability of periodontal/peri-implant soft tissue. Collagen density was also observed by fluorescent staining. As a result, surface roughness (Sa) of Ms and Rs was 0.16 μm and 0.25 μm, respectively. Protein adsorption capability on both surface showed no significant differences. In the NT group of the rat implant model, presence of HRP was restricted only in the coronal portion of epithelium. In both implant groups, in contrast, more invasion of HRP was observed in the soft tissue around implants. Especially in the Ms group, more HRP was observed in the deeper area compared with Rs group. Stronger expression of collagen was observed around Rs compared to Ms at the connective tissue-implant interface. It could be speculated that, with dense collagen, Rs implants showed stronger soft tissue integration compared with Ms implants, but the integration is not as strong as NT's.