Maximum Rubisco activities (V(cmax)), rates of photosynthetic electron transport (J(max)), and leaf nitrogen and chlorophyll concentrations were studied along a light gradient in the canopies of four temperate deciduous species differing in shade tolerance according to the ranking: Populus tremula L. < Fraxinus excelsior L. < Tilia cordata Mill. = Corylus avellana L. Long-term light environment at the canopy sampling locations was characterized by the fractional penetration of irradiance in the photosynthetically active spectral region (I(sum)). We used a process-based model to distinguish among photosynthesis limitations resulting from variability in fractional nitrogen investments in Rubisco (P(R)), bioenergetics (P(B), N in rate-limiting proteins of photosynthetic electron transport) and light harvesting machinery (P(L), N in chlorophyll and thylakoid chlorophyll-protein complexes). On an area basis, V(cmax) and J(max) (V(a) (cmax) and J(a) (max)) increased with increasing growth irradiance in all species, and the span of variation within species ranged from two (T. cordata) to ten times (C. avellana). Examination of mass-based V(cmax) and J(max) (V(m) (cmax) and J(m) (max)) demonstrated that the positive relationships between area-based quantities and relative irradiance mostly resulted from the scaling of leaf dry mass per area (M(A)) with irradiance. Although V(m) (cmax) and J(m) (max) were positively related to growth irradiance in C. avellana, and J(m) (max) was positively related to irradiance in P. tremula, the variation range was only a factor of two. Moreover, V(m) (cmax) and J(m) (max) were negatively correlated with relative irradiance in T. cordata. Rubisco activity in crude leaf extracts generally paralleled the gas-exchange data, but it was independent of light in T. cordata, suggesting that declining V(m) (cmax) with increasing relative irradiance was related to increasing diffusive resistances from the intercellular air spaces to the sites of carboxylation in this species. Because irradiance had little effect on foliar nitrogen concentration, the relationships of P(B) and P(R) with irradiance were similar to those of V(m) (cmax) and J(m) (max). Shade-intolerant species tended to have greater P(B) and P(R) and also larger V(a) (cmax) and J(a) (max) than more shade-tolerant species. However, for the whole material, P(B) and P(R) varied only about 50%, whereas V(a) (cmax) and J(a) (max) varied more than 15-fold, further emphasizing the importance of leaf anatomical plasticity in determining photosynthetic acclimation to high irradiance. Leaf chlorophyll concentrations and fractional nitrogen investments in light harvesting increased hyperbolically with decreasing irradiance to improve quantum use efficiency for incident irradiance. The effect of irradiance on P(L) was of the same order as its effect in the opposite direction on M(A), leading to either a constant model estimate of leaf absorptance with I(sum) or a slightly positive correlation. We conclude that leaf morphological plasticity is a more relevant determinant of foliage adaptation to high irradiance than foliage biochemical properties, whereas biochemical adaptation to low irradiance is of the same magnitude as the anatomical adjustments. Although shade-tolerant species did not have greater chlorophyll concentrations and P(L) than shade-intolerant species, they possessed lower M(A), and could maintain a more extensive foliar display for light capture with constant biomass investment in leaves.