Background and objective Obstructive sleep apnoea (OSA) and dyslipidaemia are independent risk factors for cardiovascular disease. This study investigates the association between OSA and plasma lipid concentrations in patients enrolled in the European Sleep Apnea Database (ESADA) cohort. Methods The cross-sectional analysis included 8592 patients without physician-diagnosed hyperlipidaemia or reported intake of a lipid-lowering drug (age 50.1 +/- 12.7 years, 69.1% male, BMI: 30.8 +/- 6.6 kg/m(2), mean apnoea-hypopnoea index (AHI): 25.7 +/- 25.9 events/h). The independent relationship between measures of OSA (AHI, oxygen desaturation index (ODI), mean and lowest oxygen saturation) and lipid profile (total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and fasting triglycerides (TG)) was determined by means of general linear model analysis. Results There was a dose response relationship between TC and ODI (mean +/- SE (mg/dL): 180.33 +/- 2.46, 184.59 +/- 2.42, 185.44 +/- 2.42 and 185.73 +/- 2.44; P < 0.001 across ODI quartiles I-IV). TG and LDL concentrations were better predicted by AHI than by ODI. HDL-C was significantly reduced in the highest AHI quartile (mean +/- SE (mg/dL): 48.8 +/- 1.49 vs 46.50 +/- 1.48; P = 0.002, AHI quartile I vs IV). Morbid obesity was associated with lower TC and higher HDL-C values. Lipid status was influenced by geographical location with the highest TC concentration recorded in Northern Europe. Conclusion OSA severity was independently associated with cholesterol and TG concentrations.