INTRODUCTION: Access to custom 3D printed pelvic implants (3DPI) is improving for application in both arthroplasty revision and tumour reconstruction. There is limited evidence regarding the safety and outcomes of such implants for large bony defects of the pelvis. The aim of this study is to report the incidence of complications, patient mortality and implant survival following pelvic reconstruction using custom 3Dprinted prostheses in the setting of extensive pelvic bone defects following pelvic tumour resection or failure of total hip arthroplasty (THA). MATERIALS AND METHODS: Patients who underwent reconstruction with a custom 3D printed pelvic prosthesis (3DPI) were identified from our clinical outcomes registry (Complex Reconstruction and Sarcoma Surgical Outcomes Registry; ANZCTRN 12621001421820). Indications for surgery, adverse events, reoperations and rates and modes of failure were recorded. Kaplan-Meier and multistate survival curves were generated for cumulative survival based on indication. RESULTS: One hundred and six procedures were completed (Revision THA = 33; Tumour Pelvis = 73) with a median follow up of 4.1 years, ranging from 0.6 to 10 years. Acetabular loosening was the most frequent indication for the Revision THA cohort, while indications for tumour varied across primary presentations, metastases and failures of previous resection/reconstruction. Intraoperative complications were observed in 4.1% (95%CI 1.1-12.3) of Tumour Pelvis cases. Overall implant retention was 96% (90-99). No mortality events were observed in the Revision THA cohort, with 5-year patient survival 79% (70-90) in the Tumour Pelvis cohort. Procedure-survival free from periprosthetic infection was 86% (74-100) in the Revision THA cohort and 85% (76-95) in the Tumour Pelvis cohort. Modelling adverse events using multistate survival models in both cohorts revealed complex time-varying presentation of adverse events, with a significant burden of reoperations and local tumour recurrence in the Tumour Pelvis cohort. CONCLUSIONS: 3DPIs are a safe and viable option for complex reconstruction of the pelvis across a range of oncological and non-oncological indications. The initial results of the present study provide important information to aid in counselling patients about such procedures and allocating healthcare resources for ongoing care. Further work is required to document functional and biomechanical outcomes in these patient populations. TRIAL REGISTRATION: Australian New Zealand Clinical Trials Registry (ANZCTRN 12621001421820); Registration date 21-Oct-2021.