Purpose: Passive fit of implant-supported restorations is difficult to achieve. The aim of this randomized clinical trial was to test the null hypotheses that a reduction in misfit strain does not occur over time, regardless of the initial strain level, and that changes in prosthesis-induced bone strain do not differ between restorations with two different levels of fit. Materials and Methods: Twenty edentulous sites were restored with screw-retained fixed restorations on two implants either cast in one piece (misfit) or assembled by an intraoral bonding procedure (fit). The restorations had a bar-shaped pontic onto which a strain-gauged metal plate could be fixed. Repeated strain gauge measurements on patient-specific in vitro resin models and on the implants intraorally were performed every 4 weeks for a period of 6 months. Statistical analysis was based on Kruskal-Wallis tests, t tests, Welch two-sample t test (α = .05), and linear regression analysis. Results: No reduction in misfit strain could be observed in vitro, indicating that alterations in prosthetic fit had not occurred in any restoration (Kruskal-Wallis; P > .05). Initial strain levels varied widely and differed significantly between fit and misfit restorations (t test; P = .0032). Regrouping the restorations with respect to a threshold strain level of 100 μm/m resulted in 10 fit and 9 misfit restorations; 1 restoration had to be excluded from analysis due to a malfunctioning strain gauge. Seven fit restorations and five misfit restorations showed strain reduction in vivo based on linear regression, while percentage strain reduction did not differ between groups (Welch two-sample t test; P = .8186). Conclusion: Within the limitations of this randomized clinical trial characterized by an observation period of 6 months and only healthy subjects being enrolled, bone adaptation around statically and dynamically loaded implants occurred, causing a decrease in misfit strain evoked by non-passively fitting prostheses. For maintaining osseointegration of dental implants, passivity of fit of multiunit restorations seems not to be as critical as previously thought. Palabras clave: bone adaptation, misfit stress, passive fit, randomized clinical trial, strain gauge