With their 2007 classification - shift work involving "circadian disruption" is probably carcinogenic to humans (Group 2A) - the International Agency for Research on Cancer [IARC] provided a riddle for scientists and the public alike. Thereafter, eighteen epidemiological investigations into shift work and a host of malignant endpoints (including cancers of the breast, prostate, lung, colon, rectum, pancreas, bladder, skin and non-Hodgkin lymphoma [NHL]) as well as mortality were published. Although IARC experts identified "circadian disruption" as the critical link in the "probable" chains of cancer causation, almost none of the post-IARC studies specifically considered a disturbed temporal organization of biology. This implies that epidemiological research to-date is less focused than it should be. To illustrate a potential chronobiological fallacy of past studies, we offer a thought experiment. In addition, we consider first empirical evidence from recent research which avoided such bias. Methodological perspectives for future chronobiology-driven epidemiological research are outlined.