ObjectivesAir pollution has been associated with increased mortality and morbidity in several studies with indications that its effect could be more severe in children. This study examined the relationship between short-term variations in criteria air pollutants and occurrence of sudden infant death syndrome (SIDS).DesignWe used a case-crossover study design which is widely applied in air pollution studies and particularly useful for estimating the risk of a rare acute outcome associated with short-term exposure.SettingThe study used data from the West Midlands region in the UK.ParticipantsWe obtained daily time series data on SIDS mortality (ICD-9: 798.0 or ICD-10: R95) for the period 1996–2006 with a total of 211 SIDS events.Primary outcome measuresDaily counts of SIDS events.ResultsFor an IQR increase in previous day pollutant concentration, the percentage increases (95% CI) in SIDS were 16 (6 to 27) for PM₁₀, 1 (−7 to 10) for SO₂, 5 (−4 to 14) for CO, −17 (−27 to –6) for O₃, 16 (2 to 31) for NO₂and 2 (−3 to 8) for NO after controlling for average temperature and national holidays. PM₁₀and NO₂showed relatively consistent association which persisted across different lag structures and after adjusting for copollutants.ConclusionsThe results indicated ambient air pollutants, particularly PM₁₀and NO₂, may show an association with increased SIDS mortality. Thus, future studies are recommended to understand possible mechanistic explanations on the role of air pollution on SIDS incidence and the ways in which we might reduce pollution exposure among infants.