Between 1988 and 1997, 18 irradiated patients (group 1, 83 implants) and 22 nonirradiated patients (group 2, 92 implants) received resection of the cancer-involved mandible and floor of the mouth and subsequently underwent mandibular rehabilitation with endosseous implants. Implant-supported prostheses were placed in 26 patients, while 13 patients received implant-tissue-supported prostheses. Between 1988 and 1991, patients were restored with implant-tissue-supported prostheses (based on 2 to 4 implants). This strategy was later changed because of the development of denture-related lesions. Since 1992, group 1 patients have been restored exclusively with implant-supported prostheses on 5 to 6 implants; group 2 patients have been rehabilitated alternatively with implant-tissue-supported prostheses on 4 implants. Special criteria for determining the success of implant-supported maxillofacial prostheses were developed. With a mean follow-up period of 37 months, 160 implants (91%) were clinically osseointegrated. Both types of restorations provided sufficient oral rehabilitation. However, only completely implant-supported prostheses avoided soft tissue ulcers. The cumulative success rate was approximately 75% after 7 years for group 1 patients and about 86% after 10 years for group 2 patients. The success rates for implants placed after the change in strategy were approximately 86% (group 1) and 94% (group 2) after 5 years. Based on these experiences, it is suggested that irradiated patients should be restored with exclusively implant-supported prostheses, without any mucosal contact. Keywords: dental implants, irradiation, maxillofacial prosthetics, oral cancer patients, success criteria