A prothrombotic state may contribute to the elevated cardiovascular risk in patients with obstructive sleep apnea (OSA). We investigated the relationship between apnea severity and hemostasis factors and effect of continuous positive airway pressure (CPAP) treatment on hemostatic activity. We performed full overnight polysomnography in 44 OSA patients (mean age 47+/-10 years), yielding apnea-hypopnea index (AHI) and mean nighttime oxyhemoglobin saturation (SpO2) as indices of apnea severity. For treatment, subjects were double-blind randomized to 2 weeks of either therapeutic CPAP (n = 18), 3 l/min supplemental nocturnal oxygen (n = 16) or placebo-CPAP (<1 cm H2O) (n = 10). Levels of von Willebrand factor antigen (VWF:Ag), soluble tissue factor (sTF), D-dimer, and plasminogen activator inhibitor (PAI)-1 antigen were measured in plasma pre- and posttreatment. Before treatment, PAI-1 was significantly correlated with AHI (r = 0.47, p = 0.001) and mean nighttime SpO2 (r = -0.32, p = 0.035), but these OSA measures were not significantly related with VWF:Ag, sTF, and D-dimer. AHI was a significant predictor of PAI-1 (R2 = 0.219, standardized beta = 0.47, p = 0.001), independent of mean nighttime SpO2, body mass index (BMI), and age. A weak time-by-treatment interaction for PAI-1 was observed (p = 0.041), even after adjusting for age, BMI, pre-treatment AHI, and mean SpO2 (p = 0.046). Post hoc analyses suggested that only CPAP treatment was associated with a decrease in PAI-1 (p = 0.039); there were no changes in VWF:Ag, sTF, and D-dimer associated with treatment with placebo-CPAP or with nocturnal oxygen. Apnea severity may be associated with impairment in the fibrinolytic capacity. To the extent that our sample size was limited, the observation that CPAP treatment led to a decrease in PAI-1 in OSA must be regarded as tentative.