This paper aimed at identifying gaps in the evidence for work stress as a risk factor for coronary heart disease (CHD) and at providing ideas for more rigorous tests of the association. Evidence in this field is mixed. The risk of type I and II errors would be reduced in future studies if work stress were assessed with predetermined standard instruments repeated over time, if outcomes excluded diagnoses based on subjective symptoms, and if individual participant data from multiple study populations were pooled to allow well-powered subgroup analyses and detailed assessments of the shape of the association. Within the Mendelian randomization and gene × environment interaction frameworks, there may be potential for using genetic data to reduce the risk of confounding and bias and to explicate the biological mechanisms underlying the association between work stress and CHD. If the evidence converges, large-scale intervention studies would be indicated despite the extensive practical problems associated with them.