Drug combination studies have become increasingly important, especially in treating cancers. Researchers are interested in identifying compounds that act synergistically when combined. However, synergy can be defined in different ways, depending on assumptions regarding drug interaction mechanisms and also depending on the different stages of drug development. Using in vitro studies, more drug doses can be studied in combination and refined statistical methods can be applied. For example, Loewe additivity and Bliss independence are two widely used methods in such studies. In general, there are fewer doses tested in combination when the studies are carried out in vivo when compared with the number tested in vitro. We found that synergy is often not clearly defined at the in vivo stage and the correct statistical analysis is not considered.