ABSTRACTIn this study it has been shown that increased diffusional resistances caused by salt stress may be fully overcome by exposing attached leaves to very low [CO2] (∼ 50 µmol mol−1), and, thus a non‐destructive‐in vivo method to correctly estimate photosynthetic capacity in stressed plants is reported. Diffusional (i.e. stomatal conductance, gs, and mesophyll conductance to CO2, gm) and biochemical limitations to photosynthesis (A) were measured in two 1‐year‐old Greek olive cultivars (Chalkidikis and Kerkiras) subjected to salt stress by adding 200 mm NaCl to the irrigation water. Two sets of A–Ci curves were measured. A first set of standard A–Ci curves (i.e. without pre‐conditioning plants at low [CO2]), were generated for salt‐stressed plants. A second set of A–Ci curves were measured, on both control and salt‐stressed plants, after pre‐conditioning leaves at [CO2] of ∼ 50 µmol mol−1 for about 1.5 h to force stomatal opening. This forced stomata to be wide open, and gs increased to similar values in control and salt‐stressed plants of both cultivars. After gs had approached the maximum value, the A–Ci response was again measured. The analysis of the photosynthetic capacity of the salt‐stressed plants based on the standard A–Ci curves, showed low values of the Jmax (maximum rate of electron transport) to Vcmax (RuBP‐saturated rate of Rubisco) ratio (1.06), that would implicate a reduced rate of RuBP regeneration, and, thus, a metabolic impairment. However, the analysis of the A–Ci curves made on pre‐conditioned leaves, showed that the estimates of the photosynthetic capacity parameters were much higher than in the standard A–Ci responses. Moreover, these values were similar in magnitude to the average values reported by Wullschleger (Journal of Experimental Botany 44, 907–920, 1993) in a survey of 109 C3 species. These findings clearly indicates that: (1) salt stress did affect gs and gm but not the biochemical capacity to assimilate CO2 and therefore, in these conditions, the sum of the diffusional resistances set the limit to photosynthesis rates; (2) there was a linear relationship (r2 = 0.68) between gm and gs, and, thus, changes of gm can be as fast as those of gs; (3) the estimates of photosynthetic capacity based on A–Ci curves made without removing diffusional limitations are artificially low and lead to incorrect interpretations of the actual limitations of photosynthesis; and (4) the analysis of the photosynthetic properties in terms of stomatal and non‐stomatal limitations should be replaced by the analysis of diffusional and non‐diffusional limitations of photosynthesis. Finally, the C3 photosynthesis model parameterization using in vitro‐measured and in vivo‐measured kinetics parameters was compared. Applying the in vivo‐measured Rubisco kinetics parameters resulted in a better parameterization of the photosynthesis model.