Background Early onset of menopause is a risk factor for several health problems. The objective was primarily to investigate the association between early menopause and current, past active and passive smoking. A second aim was to investigate the association between coffee and alcohol consumption and early menopause. Methods The present population-based cross-sectional study included a sub-sample of 2123 postmenopausal women born in 1940–41 who participated in the Oslo Health Study. Early menopause was defined as menopause occurring at an age of less than 45 years. We applied logistic regression analyses (crude and adjusted odds ratio (OR)) to examine the association between early menopause and selected lifestyle factors. Results Current smoking was significantly associated with early menopause (adj. OR, 1.59; 95% CI, 1.11–2.28). Stopping smoking more than 10 years before menopause considerably reduced the risk of early menopause (adj. OR, 0.13; 95% CI, 0.05–0.33). Total exposure to smoking (the product of number of cigarettes per day and time as a smoker) was positively related to early menopause and, at the highest doses, nearly doubled the odds (adj. OR, 1.93; 95% CI, 1.12–3.30). These data suggest a possible dose-response relationship between total exposure to smoking and early menopause, but no dose-response relationship was detected for the other variables examined. We found no significant association of coffee or alcohol consumption with early menopause. Of the lifestyle factors tested, high educational level (adj. OR, 0.50; 95% CI, 0.34–0.72) and high social participation (adj. OR, 0.60, 95% CI, 0.39–0.98) were negatively associated with early menopause. Conclusion This cross-sectional study shows an association between current smoking and early menopause. The data also suggest that the earlier a woman stops smoking the more protected she is from early menopause. Early menopause was not significantly associated with passive smoking, or alcohol or coffee consumption.