Sixty-seven patients ranging in age from 16 to 86 years were included in this 3- to 5-year retrospective report focusing on implant survival and marginal bone remodeling in relation to implant diameter. A total of 299 Brånemark implants (3.75-mm diameter: 141; 4.0-mm diameter: 61; 5.0-mm diameter: 97) were placed in 16 completely and 51 partially edentulous arches. Seven of the 141 implants in the 3.75-mm-diameter group failed (5%). The corresponding value for the 4.0-mm-wide implants was 2 of 61 (3%). The highest failure rate, 18% (17/97), was seen for the 5.0-mm-diameter implants. The least favorable cumulative survival rates were seen in mandibles after 5 years and involving 4.0-mm- and 5.0-mm-diameter implants (84.8% and 73.0%, respectively). The marginal bone loss was generally low over the 5-year period. When the data were evaluated by the Cox regression analysis, a relationship was found between implant failure and implant diameter (P .05), with a higher failure rate for the 5.0-mm-diameter implant. However, no relationship could be seen between implant failure and jaw type, or bone quality and quantity (P > .05). Neither was any relationship seen between marginal bone loss and bone quality and quantity, implant diameter, or jaw type when tested by multiple linear regression analysis (P > .05). A learning curve, poor bone quality, and changed implant design were suggested as possible reasons for the less positive outcome seen for the 5.0-mm-diameter implant. The fact that this implant was often used as a rescue implant when the standard ones were not considered suitable or did not reach initial stability was another plausible explanation. Parole chiave: Brånemark system, diameter, titanium oral implant, treatment result