Tobacco (Nicotiana tabacum L.) plants transformed with ‘antisense’ rbcS to decrease the expression of ribulose-1,5-bisphosphate carboxylase-oxygenase (Rubisco) have been used to investigate the contribution of Rubisco to the control of photosynthesis in plants growing at different irradiances. Tobacco plants were grown in controlled-climate chambers under ambient CO2 at 20°C at 100, 300 and 750 μmol·m−2·s−1 irradiance, and at 28°C at 100, 300 and 1000 μmol·m−2·s−1 irradiance. (i) Measurement of photosynthesis under ambient conditions showed that the flux control coefficient of Rubisco (C infRubisco supA ) was very low (0.01–0.03) at low growth irradiance, and still fairly low (0.24–0.27) at higher irradiance. (ii) Short-term changes in the irradiance used to measure photosynthesis showed that C infRubisco supA increases as incident irradiance rises, (iii) When low-light (100 μmol·m−2·s−1)-grown plants are exposed to high (750–1000 μmol·m−2·s−1) irradiance, Rubisco is almost totally limiting for photosynthesis in wild types. However, when high-light-grown leaves (750–1000 μmol·m−2·s−1) are suddenly exposed to high and saturating irradiance (1500–2000 μmol·m−2·s−1), C infRubisco supA remained relatively low (0.23–0.33), showing that in saturating light Rubisco only exerts partial control over the light-saturated rate of photosynthesis in “sun” leaves; apparently additional factors are co-limiting photosynthetic performance, (iv) Growth of plants at high irradiance led to a small decrease in the percentage of total protein found in the insoluble (thylakoid fraction), and a decrease of chlorophyll, relative to protein or structural leaf dry weight. As a consequence of this change, high-irradiance-grown leaves illuminated at growth irradiance avoided an inbalance between the “light” reactions and Rubisco; this was shown by the low value of C infRubisco supA (see above) and by measurements showing that non-photochemical quenching was low, photochemical quenching high, and NADP-malate dehydrogenase activation was low at the growth irradiance. In contrast, when a leaf adapted to low irradiance was illuminated at a higher irradiance, Rubisco exerted more control, non-photochemical quenching was higher, photochemical quenching was lower, and NADP-malate dehydrogenase activation was higher than in a leaf which had grown at that irradiance. We conclude that changes in leaf composition allow the leaf to avoid a one-sided limitation by Rubisco and, hence, overexcitation and overreduction of the thylakoids in high-irradiance growth conditions, (v) ‘Antisense’ plants with less Rubisco contained a higher content of insoluble (thylakoid) protein and chlorophyll, compared to total protein or structural leaf dry weight. They also showed a higher rate of photosynthesis than the wild type, when measured at an irradiance below that at which the plant had grown. We propose that N-allocation in low light is not optimal in tobacco and that genetic manipulation to decrease Rubisco may, in some circumstances, increase photosynthetic performance in low light.