Wheat (Triticum aestivum L.) cv. Minaret was grown in open-top chambers (OTCs) in 1995 and 1996 under three carbon dioxide (CO2) and two ozone (O3) levels. Plants were harvested regularly between anthesis and maturity to examine the rate of grain growth (dG/dt; mg d–1) and the rate of increase in harvest index (dHI/dt;% d–1). The duration of grain filling was not affected by elevated CO2 or O3, but was 12 days shorter in 1995, when the daily mean temperature was over 3 °C higher than in 1996. Season-long exposure to elevated CO2 (680 μmol mol–1) significantly increased the rate of grain growth in both years and mean grain weight at maturity (MGW) was up to 11% higher than in the chambered ambient air control (chAA; 383 μmol mol–1). However, the increase in final yield obtained under elevated CO2 relative to the chAA control in 1996 resulted primarily from a 27% increase in grain number per unit ground area. dG/dt was significantly reduced by elevated O3 under ambient CO2 conditions in 1995, but final grain yield was not affected because of a concurrent increase in grain number. Neither dG/dt nor dHI/dt were affected by the higher mean O3 concentrations applied in 1996 (77 vs. 66 nmol mol–1); the differing effects of O3 on grain growth in 1995 and 1996 observed in both the ambient and elevated CO2 treatments may reflect the contrasting temperature environments experienced. Grain yield was nevetheless reduced under elevated O3 in 1996, primarily because of a substantial decrease in grain number. The data obtained show that, although exposure to elevated CO2 and O3 individually or in combination may affect both dG/dt and dHI/dt, the presence of elevated CO2 does not protect against substantial O3-induced yield losses resulting from its direct deleterious impact on reproductive processes. The implications of these results for food production under future climatic conditions are considered.