Multiple cancer drug candidates in clinical trials kill tumor cells through off-target effects instead of by interacting with their intended molecular targets, according to a new study. The unexpected findings demonstrate that the targets of these drugs are not essential for the survival of cancer cells - contradicting over 180 previous reports about their importance - and may help explain why seemingly effective cancer drugs often fail to be translated to the clinic. Most cancer drugs that are tested in preclinical and clinical studies never receive FDA approval, largely because the drugs turn out to be too toxic or ineffective in humans. However, it is unclear why so many candidates run into these issues. Ann Lin and colleagues previously discovered that the small molecule OTS167 killed cancer cells by inhibiting proteins other than its designated target. In this study, the authors used CRISPR gene editing techniques to examine the mechanisms of ten other cancer drugs that target one of six proteins, which have been reported as important for the survival of cancer cells in over 180 publications. The drugs studied have been used in at least 29 different clinical trials involving a total of over 1,000 patients, and include prominent candidates such as citarinostat and ricolinostat, which are being tested against multiple myeloma. Contrary to the previous reports based on RNA silencing, the drugs did not actually kill cancer cells by inhibiting their target proteins: they still worked when given to cells deficient in their target. Rather, the drugs induced cell death through off-target mechanisms; for example, the authors found the true target of the drug candidate OTS964 (designed to target the PBK enzyme) was another enzyme named CDK11. Lin et al. argue that it will be necessary to adopt more rigorous genetic approaches in preclinical trials to verify that future drug candidates work as intended.