Observations of nonuniform photosynthesis across leaves cast doubt on intemal CO2 partial pressures (p) calculated on the assumption of uniformity and can lead to incorrect conclusions about the stomatal control of photosynthesis. The problem can be avoided by measuring pi directly because the assumptions of uniformity are not necessary. We therefore developed a method that allowed pi to be measured confinuously in situ for days at a time under growth conditions and used it to investigate intact leaves of sunflower (Helianthus annuus L.), soybean (Glycine max L. Merr.), and bush bean (Phaseolus vulgaris L.) subjected to high or low leaf water potentials (I,) or high concentrations of abscisic acid (ABA). The leaves maintained a relatively constant differential (Ap) between ambient CO2 and measured pi throughout the light period when water was supplied. When water was withheld, ', decreased and the stomata began to close, but measured p increased until the leaf reached a *w of -1.76 (bush bean), -2.12 (sunflower) or -3.10 (soybean) megapascals, at which point Ap = 0. The increasing pi indicated that stomata did not inhibit CO2 uptake and a Ap of zero indicated that CO2 uptake became zero despite the high availability of CO2 inside the leaf. In contrast, when sunflower leaves at high Iw were treated with ABA, pi did not increase and instead decreased rapidly and steadily for up to 8 hours even as I, increased, as expected if ABA treatment primarily affected stomatal conductance. The accumulating CO2 at low Iw and contrasting response to ABA indicates that photosynthetic biochemistry limited photosynthesis at low Iw but not at high ABA.Water deficits have a large impact on plant growth and productivity by reducing leaf turgor, cell expansion, stomatal conductance, and photosynthesis and increasing ABA and solute concentrations in the tissues (2, 4). Among these responses, the inhibition of photosynthesis is central and often thought to be caused by reduced stomatal conductance, perhaps because of high ABA, that limits the availability of CO2 inside the leaf (8-10, 29, 32, 33). However, the '