Introduction Consistent evidence has shown a positive association between air pollution and daily mortality among adults. Less is known about its effect on infant mortality and the modification of this association by socioeconomic status (SES).

Objective To assess the association of particulate matter with an aerodynamic diameter of ≤10 μm (PM₁₀) and ozone (O₃) on infant mortality and its modification by SES.

Methods We evaluated the relationship of 24 h mean PM₁₀ and 1 h daily maximum O₃ levels with 12 079 all-cause deaths (3903 respiratory deaths) among 1- to 11-month-old infants residing in the Mexico City Metropolitan Area between January 1997 and December 2005 using a case-crossover approach. The data were analysed using conditional logistic regression models, controlling for weather conditions and day of the week.

Results Single-effect models showed, for all-cause mortality, increases of 5.5% (95% CI 1% to 10%) at lag1 and 6.6% (2% to 11.4%) at lag2; cumulative exposure models (0–2 days) showed an increase of 6.3% (0.01% to 32.7%). Respiratory mortality increased marginally at 5.3% (−0.02% to 13.2%) with a 1-day lag and 10% (2.1% to 18%) with a 2-day lag per increase of 38.7 μg/m³ (IQR) in PM₁₀ levels. When data were stratified by SES (low, medium, and high), only infants with low and medium SES presented a significant increase in risk of all-cause mortality and respiratory mortality in relation to PM₁₀. O₃ was only significantly related to respiratory mortality in low SES.

Conclusion Our results suggest that in the Mexico City Metropolitan Area, infants with lower SES (low to medium) are at higher risk of mortality when exposed to ambient PM₁₀ and O₃.