Drug combinations have been proposed to combat drug resistance, but putative treatments are challenged by low bench-to-bed translational efficiency. To explore the effect of cell culture format and readout methods on identification of synergistic drug combinations in vitro, we studied response to 21 clinically relevant drug combinations in standard planar (2D) layouts and physiologically more relevant spheroid (3D) cultures of HCT-116, HT-29 and SW-620 cells. By assessing changes in viability, confluency and spheroid size, we were able to identify readout-and culture format-independent synergies, as well as synergies specific to either culture format or readout method. In particular, we found that spheroids, compared to 2D cultures, were generally both more sensitive and showed greater synergistic response to combinations involving a MEK inhibitor. These results further shed light on the importance of including more complex culture models in order to increase the efficiency of drug discovery pipelines. Colorectal cancer (CRC) is the third most common neoplastic malignancy worldwide 1 , and although improvements in standard treatments have increased the survival rates over the past 20 years 2 , far from all patients benefit from currently available therapies. Targeted therapy, using drugs aimed to target specific molecules involved in tumour growth, is being regarded as a promising tool to increase response rates to cancer therapy. However, the number of such therapies that have made it all the way to the clinic has been limited. This may be explained by lack of therapy response due to adaptive drug resistance, or transient response due to acquired resistance. Drug combinations are being discussed as a promising strategy to overcome the resistance frequently observed upon administration of targeted monotherapy 3,4. The augmented effect of targeted drug combination treatment is frequently ascribed to the drugs' ability to jointly interfere with the growth-promoting signalling network of cancer cells at multiple points. High-throughput cell line screening platforms have been successfully employed as tools to uncover novel synergistic drug combinations. In the study ALMANAC of the National Cancer Institute (NCI), where a large number of pairwise combinations of FDA-approved cancer drugs were screened in vitro, several novel pairs of synergistic drug combinations were identified, whereof roughly a third also were shown to be efficient and synergistic in vivo 5. Another example is the Merck Research Laboratories screen, in which 583 combinations of experimental and approved cancer drugs were screened in a panel of cancer cell lines, identifying well-known as well as novel synergistic drug combinations in vitro 6. Despite large combination screening efforts with successful hits in vitro, putative treatments are challenged by low bench-to-bed translational efficiency. The insufficient ability of cell lines grown on planar surfaces to correctly recapitulate drug response in vivo has been debated as a possible explanation ...