The association between excess cortisol and various parameters of metabolic syndrome including hypertension, insulin resistance and dyslipidaemia is increasingly recognised. The present single-blind randomised placebo-controlled cross-over study compared the effect of polyphenol-rich dark chocolate (DC) on biomarkers of glucose metabolism, lipid profile, and blood pressure (BP) in females with BMI ≥ 25 kg m−2 (n = 21) and females with BMI < 25 kg m−2 (n = 21). Volunteers consumed 20 g of DC containing 500 mg polyphenols or a placebo DC with negligible polyphenol-content daily for 4 weeks, separated by a 2-week washout period. Systolic BP and diastolic BP decreased after 4 weeks of polyphenol-rich DC. Placebo raised fasting insulin, homeostasis model assessment of insulin resistance (HOMA-IR) and salivary cortisol, an effect that was significantly different from polyphenol-rich DC which had a negligible effect on fasting insulin, HOMA-IR and salivary cortisol. Females with BMI ≥ 25 kg m−2 responded less favourably to placebo than lean females and consequently had higher fasting insulin and HOMA-IR, in addition to a lower quantitative sensitivity check index (QUICKI) after ingestion of placebo compared to polyphenol-rich DC. No significant changes in lipid profile were observed. This study provides evidence for the metabolic benefits of consuming polyphenol-rich dark chocolate while demonstrating the possibility of adverse effects occurring with polyphenol-poor chocolate placebo.