Background: Studies on long-term exposure to air pollution at high levels suggest an association with increased risk of metabolic syndrome (MetS). We aimed to study this relationship at very low levels, less well-studied, yet of public health importance. Methods: The Swedish CArdioPulmonary bioImage Study (SCAPIS) consists of randomly selected subjects aged 50–64 years from six large cities in Sweden (n = 30,154). Participants underwent medical examination and answered a health-related questionnaire at enrolment. MetS was defined using established criteria. Residential exposure to PM2.5 and NO2 was modelled annually for each subject over 10 years before enrolment. Adjusted prevalence ratios (adj PR) were calculated both by quartiles of exposure and per 1 μg/m3 increment in PM2.5 and NO2, with adjustments for age, lifestyle and socioeconomic factors. Results: Full information on exposure, covariates, and MetS was obtained for 13,997 (90.26 %) females and 12,978 (88.61 %) males, respectively. The 10-years mean exposures of PM2.5 and NO2 before enrolment were 6.57 and 11.47 μg/m3, respectively. The prevalence of MetS was 28.75 % in females and 38.20 % in males, respectively. Quartile analysis of PM2.5 did not reveal significant increased adj PR for MetS, neither in females nor in males. For females the adj PR per 1 μg/m3 was 0.91 (0.89–0.94) and for males 0.95 (0.92–0.97), respectively. Similar, but somewhat weaker, associations were seen for NO2. Conclusion: Some inverse associations between air pollution and MetS were observed, but these results should be interpreted with caution due to limitations in the cross-sectional design. Results should not be interpreted beyond our studied exposure range.