BACKGROUND: We have previously found a quite strong interplay between occupational airborne pollutants, ABO phenotypes, and risk of ischaemic heart disease (IHD), with long-term exposure being associated with a significantly increased risk among men with phenotype O, and not among men with other ABO phenotypes. We suggested that the biological pathway could be a stronger systemic inflammatory response in men with blood group O. Several inflammatory mediators likely to increase the risk of IHD have recently been linked also to obesity, suggesting that long-term exposure to airborne pollutants might play a role in the aetiology of obesity. Accordingly, we tested the hypothesis that long-term occupational exposure to airborne pollutants would be more strongly associated with obesity in men with phenotype O than in men with other ABO phenotypes. DESIGN: Cross-sectional exposure-response study taking into account potential confounders. SETTING: The Copenhagen Male Study. SUBJECTS: A total of 3290 men aged 53-74 y. MAIN OUTCOME MEASURE: Prevalence of obesity (BMI > or =30 (kg/m2)). RESULTS: Overall, no differences were found in the prevalence of obesity between men with the O phenotype (n=1399) and men with other phenotypes (n=1891), 8.6 and 9.0%. However, only among men with the O phenotype was long-term occupational exposure (at least 5 y of frequent exposure) to various respirable airborne pollutants: dust, asbestos, soldering fumes, welding fumes, organic solvents, fumes from lacquer, paint or varnish, toxic components, breath irritants, stench or strongly smelling products, and irritants (other than breath irritants or contagious components) associated with an increased prevalence of obesity. Statistically, the strongest univariate associations were found for asbestos exposure, welding fumes, and breath irritants. Odds ratios (95% confidence limits) for these factors were 3.7 (1.8-7.6), 2.7 (1.6-4.4), and 2.6 (1.5-4.4), respectively. This particular relationship of airborne exposures with obesity in men with phenotype O was supported in multivariate analysis including interaction terms and taking into account a number of potential confounders. In contrast, no gene-environment interactions with obesity were found with respect to ABO phenotypes and a number of nonrespirable exposures. CONCLUSION: The finding of a quite strong interplay between long-term exposure to airborne pollutants, ABO phenotypes, and risk of obesity may open up new possibilities for clarifying mechanisms underlying the global obesity epidemic.