PURPOSE The aim of this study was to compare implant-supported restorations placed and loaded immediately or with a delay in a longitudinal case control study. MATERIALS AND METHODS Seventy-six patients with 222 implants were enrolled in this study. One hundred eleven implants (45 patients) were submitted to immediate functional or nonfunctional loading. These were compared to 111 implants (51 patients) that received delayed loading after submerged healing. The mean observation time was 40.3 months (3.36 years). Implant success was determined, and peri-implant soft tissue parameters and esthetic outcomes for anterior restorations were evaluated. The implants were divided into four groups according to their treatment protocol: immediate (I) or delayed (D) implant placement (P) or function (F), ie: group 1 = IF+IP, group 2 = IF+DP, group 3 = DF+IP, and group 4 = DF+DP. RESULTS Five implants were lost during healing, giving an overall success rate of 97.7%. Implants with delayed function showed significantly better results (100.0%) than implants that were immediately loaded (95.5%). Four of the five lost implants had been placed immediately postextraction (success rate for delayed implant placement, 99.4%, versus 93.1% for immediate implant placement). Regarding the four treatment protocols, group 1 showed a success rate of 91.3%; group 2 achieved 98.5%; and both delayed function groups showed 100% success. No statistically significant difference was seen between the four groups. Esthetically significant advantages were seen for the implants placed into immediate function. Probing depths and bleeding on probing were significantly lower in the group of implants placed into immediate function. CONCLUSIONS Implants that are loaded immediately can achieve good outcomes. However, the risk of implant loss appears to be increased in cases where immediate function is combined with immediate implant placement.