Glioblastoma (GBM) is a lethal brain tumor. Standard of care treatment comprising surgery, radiation and chemotherapy results in median survival rates of 12-15 months. Molecular targeted agents identified using conventional two-dimensional (2D) in vitro models of GBM have failed to improve outcome in patients, rendering such models inadequate for therapeutic target identification. A previously developed 3D GBM in vitro model that recapitulates key GBM clinical features and responses to molecular therapies was investigated for utility for screening novel radiation-drug combinations using gold-standard clonogenic survival as readout. Patient-derived GBM cell lines were optimized for inclusion in a 96-well plate 3D clonogenic screening platform, ClonoScreen3D. Radiation responses of GBM cells in this system were highly reproducible and comparable to those observed in low-throughout 3D assays. The screen methodology provided quantification of candidate drug single agent activity (EC ) and the interaction between drug and radiation (radiation interaction ratio, RIR). The PARP inhibitors talazoparib, rucaparib and olaparib, each showed a significant interaction with radiation by ClonoScreen3D and were subsequently confirmed as true radiosensitizers by full clonogenic assay. Screening a panel of DNA damage response inhibitors revealed the expected propensity of these compounds to interact significantly with radiation (13/15 compounds). A second screen assessed a panel of compounds targeting pathways identified by transcriptomic analysis and demonstrated single agent activity and a previously unreported interaction with radiation of dinaciclib and cytarabine (RIR 1.28 and 1.90, respectively). These compounds were validated as radiosensitizers in full clonogenic assays (sensitizer enhancement ratio 1.47 and 1.35, respectively). The ClonoScreen3D platform was demonstrated to be a robust method to screen for single agent and radiation-drug combination activity. Using gold-standard clonogenicity, this assay is a tool for identification of radiosensitizers. We anticipate this technology will accelerate identification of novel radiation-drug combinations with genuine translational value. [Abstract copyright: Copyright © 2024 The Authors. Published by Elsevier Inc. All rights reserved.]