In vitro combination screens in cancer cell lines provide an opportunity to discover novel synergistic interactions and gain insights into the mechanism behind these synergies. Challenges in this discovery process include variability in the experimental system, variation in the level of single drug response, different reference models and methods for defining additivity, interpretation of results and follow-up strategies to translate findings to the in vivo and clinical settings. We compared different analysis methodologies, including Bliss Independence and Loewe Additivity, and two different activity definitions, illustrating how they led to different emphasis on anti-proliferative or cell killing effects. Other important practical considerations will be discussed, including quality control, data representation and visualization, and strategies for following-up on in vitro findings.