Objectives In 2007, IARC classified shift-work involving “circadian disruption” as probably carcinogenic [Group 2A]. Thereafter, 23 shift-work studies into a series of malignant endpoints provided additional epidemiologic evidence regarding chains of cancer causation, which appear biologically plausible when experimental data are considered. Since none of the studies specifically assessed “circadian disruption” or disturbed chronobiology, we ask the following: Does published research suffer from chronobiological errors? How could we assess circadian or chronodisruption [CD] in future epidemiological studies?

Method On the basis of chronobiological insights we develop a quantitative approach to assess CD for different chronotypes and for different shift-work regimens. We examine how chronobiological errors may have affected studies published so far with a focus on those which considered chronobiological information but were confined to night-work.

Results We illustrate that ignoring “internal time”, which is critically determined by an individual’s chronotype, may lead to what we propose to call “IT errors”. In addition, assessing biological effects associated with shift-work should also include associated activities and therefore temporal information beyond the nominal shift time. Such errors of truncating relevant “external time” information (activities at chronobiologically unusual times start before and do not end with the shift) we propose to coin “ET errors”. We propose methodologically how observational research may avoid potential chronobiological biases and how chronodisruption can be computed.

Conclusions Avoiding both IT and ET errors with regard to any – and not just night – shift work may be a prerequisite to assess causal links between shift-work and cancer which may remain otherwise masked.