Insomnia is a highly prevalent disorder and a state of 24 h hyperarousal is considered as a key factor of this condition. Various physiological markers of hyperarousal have been investigated, including the activity of the HPA axis. However, these studies yielded heterogenous results. The aim of this study was to qualitatively and quantitatively evaluate whether there are differences in cortisol levels, the hormonal end product of the HPA axis, between patients with insomnia and good sleeper controls (GSC). The databases PubMed, Cinahl, PsycInfo, PsycArticles and Web of Science were searched for case–control studies comparing cortisol levels in patients with insomnia and GSC. Twenty studies (449 patients with insomnia, limited to ages 18–70 not taking any medications; 357 GSC) met the inclusion criteria. For each study, standardized mean differences (SMD) were calculated as a measure of effect size. Results suggest that patients with insomnia show moderately increased cortisol levels (SMD = 0.50, 95% CI: [0.21–0.80]). Higher effect sizes were found by including only studies using blood samples in the analysis (SMD = 0.67, 95% CI: [0.15–1.18]). Furthermore, a positive, but insignificant association was found between the extent of objective sleep loss in insomnia patients and group differences in cortisol levels.