Apoptosis resistance contributes to treatment failure in colorectal cancer (CRC). New treatments that reinstate apoptosis competency have potential to improve patient outcome but require predictive biomarkers to target them to responsive patient populations. Inhibitor of apoptosis proteins (IAPs) suppress apoptosis, contributing to drug resistance; IAP antagonists such as TL32711 have therefore been developed. We developed a systems biology approach for predicting response of CRC cells to chemotherapy and TL32711 combinations in vitro and in vivo. CRC cells responded poorly to TL32711 monotherapy in vitro; however, co-treatment with 5-fluorouracil (5-FU) and oxaliplatin enhanced TL32711-induced apoptosis. Notably, cells from genetically identical populations responded highly heterogeneously, with caspases being activated both upstream and downstream of mitochondrial outer membrane permeabilisation (MOMP). These data, combined with quantities of key apoptosis regulators were sufficient to replicate in vitro cell death profiles by mathematical modelling. In vivo, apoptosis protein expression was significantly altered, and mathematical modelling for these conditions predicted higher apoptosis resistance that could nevertheless be overcome by combination of chemotherapy and TL32711. Subsequent experimental observations agreed with these predictions, and the observed effects on tumour growth inhibition correlated robustly with apoptosis competency. We therefore obtained insights into intracellular signal transduction kinetics and their population-based heterogeneities for chemotherapy/TL32711 combinations and provide proof-of-concept that mathematical modelling of apoptosis competency can simulate and predict responsiveness in vivo. Being able to predict response to IAP antagonist-based treatments on the background of cell-to-cell heterogeneities in the future might assist in improving treatment stratification approaches for these emerging apoptosis-targeting agents.