Purpose: This study aimed to conduct a radiological and clinical evaluation of patients who underwent fixed rehabilitation involving subperiosteal implants and to define not only the criteria for failure but also the criteria for survival and success. Materials and methods: Patients suffering from severe atrophy of the maxilla and/or mandible underwent immediate loading with fixed full-arch prostheses supported by custom-made 3D-printed subperiosteal implants. The evaluated outcomes included the success and survival rates of subperiosteal implant prosthetic therapy, surgical complications, and the three-dimensional positions of osteosynthesis screws and sleeves used for implant adhesion to the bone. The SIO criteria (Stability, Inflammation, and Offset) were used for evaluating the success, survival, or failure of subperiosteal implant-prosthetic rehabilitation. Success required no clinical mobility, asymptomatic soft tissues without recession, and no intraoperative deviations from virtual implant planning; while survival allowed transient micromovements, tissue recession without inflammation, and minimal post-surgical implant modifications. Persistent micromotion and implant removal due to aesthetic dissatisfaction or fabrication of a new implant were classified as failures. Non-parametric tests were applied to the datasets with a significance level of p = 0.01. Eighteen patients were followed up for two years. Eleven implants were placed in the upper jaw and 15 in the lower jaw. Temporary instability was observed in four subperiosteal implants, accounting for a survival rate of 84.6% based on the Stability criterion). Two patients suffered from peri-implant soft tissue recession at two years (Inflammation criterion). Preoperative virtual implant positions and achieved postoperative position of 189 titanium osteosynthesis screws and sleeves were analyzed: 98 buccal, 69 linguopalatal, and 22 in other areas. The overall linear sleeve misalignment was 1.66±0.99mm. Horizontal misalignment (1.46±1.03mm) was approximately three times greater than vertical misalignment (0.59±0.21mm), with a significant difference (p value=0.0004). One patient exhibited horizontal and overall misalignments exceeding the maximum threshold suitable for implant success (approximately 5 mm), resulting in 25 successful and only one surviving implant (Offset criterion). No prosthesis failed during the entire follow-up, and no mechanical or functional complications were recorded. The cumulative survival rate was 100% at the two-year follow-up. Conclusions: The proposed SIO criteria defined success, survival, and failure for subperiosteal implant rehabilitation. Transient movements or infection events were classified as survival indicators, while negative Offset results led the implant survival but not success. The present virtual design and planning approach achieved a 73.1% success rate over two years according to the SIO classification.