The evergreen holm oak Quercus ilex L. is the most representative tree in Mediterranean forests. Accurate estimation of the limiting factors of photosynthesis for Q. ilex and the prediction of ecosystem water-use efficiency by mechanistic models can be achieved only by establishing whether this species shows heterogenic stomatal aperture, and, if so, the circumstances in which this occurs. Here, we collected gas-exchange and chlorophyll fluorescence data in Q. ilex leaves from a nursery to measure the effects of stomatal oscillations on PSII quantum yield (Φ(PSII)) under water stress. Stomatal conductance (g(s)) was used as an integrative indicator of the degree of water stress. Images of chlorophyll fluorescence showed heterogeneous Φ(PSII) when g(s) was <50 mmol H(2)O m(-2) s(-1), representative of severe drought and corresponding to a container capacity <45%. Stomatal patchiness was related to a coefficient of variation (CV) of Φ(PSII) values >2.5%. A parallel study in the forest confirmed heterogeneous Φ(PSII) values in leaves in response to declining water availability. Three kinds of Q. ilex individuals were distinguished: those resprouting after a clear-cut (resprouts, R); intact individuals growing in the same clear-cut area as resprouts (controls, C); and intact individuals in a nearby, undisturbed area (forest controls, CF). Patchiness increased in C and CF in response to increasing drought from early May to late July, whereas in R, Φ(PSII) values were maintained as a result of their improved water relations since the pre-existing roots were associated with a smaller aerial biomass. Patchiness was related to a % CV of Φ(PSII) values >4 and associated in the summer with mean g(s) values of 30 mmol H(2)O m(-2) s(-1). Under milder drought in spring, Φ(PSII) patchiness was less strictly related to g(s) variations, pointing to biochemical limitants of photosynthesis. The occurrence of heterogenic photosynthesis caused by patchy stomatal closure in Q. ilex during severe drought should be taken into account in ecosystem modelling in which harsher water stress conditions associated with climate change are predicted.