The CO2 transfer conductance in leaves quantifies the ease with which CO2 can diffuse from sub-stomatal cavities to sites of carboxylation within the chloroplast. The aim of this work was to test the hypothesis that the CO2 transfer conductance is proportional to the surface area of chloroplasts exposed to intercellular airspaces. We compared two genotypes, wild-type and transgenic tobacco, that had been transformed with an antisense gene directed at the mRNA of the Rubisco small subunit. Transgenic tobacco had lower rates of CO2 assimilation than wild-type but similar chlorophyll contents. Leaf anatomy was altered by growing plants in two different environments: a high daily irradiance in a growth cabinet (12 h photoperiod of 1 mmol quanta m-2 s-1) and a sunlit glasshouse. The growth cabinet gave at least twice the daily irradiance compared to the glasshouse. The CO2 transfer conductance was calculated from combined measurements of gas exchange and carbon isotope discrimination measured in 2% oxygen. Following gas exchange measurement, leaves were sampled for biochemical and anatomical measure- ment. In transgenic tobacco plants, Rubisco content was 35% of that found in the wild-type tobacco, the CO2 assimilation rate was 50% of the wild-type rate and the chlorophyll content was unaltered. While leaf mass per unit leaf area of transgenic tobacco was 82% of that of the wild-type, differences in leaf thickness and surface area of mesophyll cells exposed to intercellular airspace per unit leaf area (Smes) were small (92 and 87% of wild-type, respectively). Leaves grown in the growth cabinet under high daily irradiance were thicker (63%), had a greater Smes (41%) due to the development of thicker palisade tissue, had higher photosynthetic capacity (27%) and contained more chlorophyll (58%) and Rubisco (77%), than leaves from plants grown in the glasshouse. Irrespective of genotype or growth environment, CO2 transfer conductance varied in proportion to surface area of chloroplasts exposed to intercellular airspaces. While the method for calculating CO2 transfer conductance could not distinguish between limitations due to the gas or liquid phases, there was no reduction in CO2 transfer conductance associated with more closely packed cells, thicker leaves, nor with increasing chloroplast thickness in tobacco.