Cortisol has a pivotal role in physical and mental health, but relatively few studies have paid attention to individual differences in cortisol levels and the etiology of these differences, in particular their possible genetic basis. In this article we review the existing literature on the heritability of cortisol levels. Most of the studies, which have been carried out in genetically informative samples, lack methodological consistency with regard to frequency and timing of sample collection. The circadian rhythm in cortisol levels was often not taken into account. A power analysis shows that none of these studies used adequate sample sizes to distinguish genetic from shared environmental influences as a cause for familial aggregation. Results of a simultaneous analysis of 5 comparable twin studies suggest a heritability of 62%. Hence, we conclude that, to understand the contribution of genetic and (shared) environmental influences to variation in basal cortisol levels, future studies should be designed more rigorously with strict collection and sampling protocols, sufficient sample size and repeated measures across multiple days. © 2002 Elsevier Science Ltd. All rights reserved.