The aim of the study was to identify and quantify the reasons for the high bladder cancer rates in Turkey. We conducted a case–control study in Manisa, Turkey, in 2011. The study included 173 patients with incident, histologically confirmed bladder cancer and 282 controls who were frequency matched by age, sex and geographic area, admitted to the main hospital of Manisa for a wide range of acute diseases. Adjusted odds ratios (ORs) and the corresponding 95% confidence intervals (CIs) were derived from multiple logistic regression models. Compared with never smokers, the OR was 2.9 (95% CI 1.5–5.4) for moderate (<20 cigarettes/day) and 4.0 (95% CI 1.7–9.6) for heavy smokers. The association was stronger for unfiltered black tobacco (OR=5.4) and for longer duration of smoking (≥40 years, OR=5.3). There was a strong inverse correlation with social class indicators, with ORs of 0.2 (95% CI 0.1–0.4) for more-educated compared with less-educated individuals. There was no significant association with a group of five occupations a priori defined as being of high risk (OR=1.3), nor with farming (OR=1.2). Bladder cancer risk was directly related to the history of urinary tract infections (OR=1.9, 95% CI 1.2–3.1) but not to diabetes (OR=0.7) or kidney (OR=0.7) and prostate (OR=1.3) diseases. Tobacco is the major risk factor for bladder cancer in Manisa, being responsible for 56% of cases; urinary tract infections account for 19% of cases, whereas the role of occupational exposure is limited in this, predominantly rural, population.