Post-stroke levels of total cholesterol (TC) appear to be negatively associated with stroke mortality. Statin pretreatment might affect this association. Sex differences in the prognostic value of the lipid profile have not yet been studied. We have evaluated the impact of TC, high- and low-density lipoprotein (HDL and LDL, respectively), and triglyceride (TG) levels on the 3-month outcome after a first ischemic stroke (IS) according to sex and previous statin use. The study group consisted of a hospital-based cohort of consecutive patients with a diagnosis of first IS. Poor outcome was defined as a modified Rankin Scale (mRS) score >or=3 at 90 days. The odds ration (OR) for poor prognosis was analyzed for each sex using logistic regression models adjusted for vascular risk factors and statin pretreatment. A total of 591 patients were included in the analysis (318 men). The predictors of a 90-day poor outcome were age and initial NIH Stroke Scale (NIHSS) score in women, and age, initial NIHSS, smoking, atrial fibrillation, and thrombolytic treatment in men. In women, none of the lipids studied affected the 90-day prognosis. Men falling in the last quintile of TC [OR: 0.68 95% confidence interval (95% CI) 0.52-0.88; p = 0.004] and LDL (OR 0.74, 95% CI 0.56-0.98; p = 0.04) have better outcome than men in the first quintile. Adjusting for statin pretreatment did not change the results. The results indicated that an association between poststroke lipids and prognosis may vary by sex. In women, lipids were not associated with the outcome; in men, lower TC and LDL were associated with worse prognosis. These differences can not be explained by statin use and require further research.