Purpose: This prospective study aims to evaluate the clinical outcomes of 'All-on-four' rehabilitations in controlled human immunodeficiency virus (HIV)-positive patients. Materials and methods: Edentulous patients requiring an implant prosthetic restoration of one or both jaws were enrolled in the present study. Each patient received at least one fixed full-arch prosthesis. Four implants, immediately loaded, were placed in each jaw using the 'All-on-four' protocol. Marginal bone loss, implant and prosthetic failure, biological and mechanical complications, and serological levels (CD4 cell count, CD4/CD8 ratio, and HIV viral load) were recorded up to 7-year follow-up. Results: A total of 116 implants were placed in 24 patients, and 29 rehabilitations based on the 'All-on-four' concept were achieved. Implant failures were registered in four patients (10 of 116 implants), and the implant survival rate was 91.37%. At the 7-year radiographic evaluation, peri-implant crestal bone loss averaged 1.91 ± 1.3 mm for upright maxillary implants (n = 30 implants) and 1.79 ± 1.28 mm for tilted maxillary implants (n = 30 implants). In the mandible, mean peri-implant crestal bone loss was 1.54 ± 1.27 mm for upright implants (n = 28) and 1.5 ± 1.3 mm for tilted implants (n = 28). No statistically significant correlation was found between serological parameters and marginal bone levels at 6 months, or through 7 years of annual follow-up (P > 0.05). A statistically significant linear correlation (P 0.001) was found between early implant failure and HIV viral load. The CD4/CD8 ratio was significantly correlated with late implant failure (P = 0.009). Conclusions: Within the limitations of this prospective 7-year longitudinal study, HIV-positive patients with a stable immune system can be candidates for the 'All-on-four' treatment concept. Parole chiave: dental implant, HIV-patients, immediate loading, oral rehabilitation, tilted implant Conflict of interest statement: The authors declare there are no conflicts of interest.