The relationships between photosynthetic efficiency, non-radiative energy dissipation and carotenoid composition were studied in leaves ofLigustrum ovalifolium developed either under full sunlight or in the shade. Sun leaves contained a much greater pool of xanthophyll cycle components than shade leaves. The rate of non-radiative energy dissipation, measured as non-photochemical fluorescence quenching (NPQ), was strictly related to the deepoxidation state (DPS) of xanthophyll cycle components in both sun and shade leaves, indicating that zeaxanthin (Z) and antheraxanthin (A) are involved in the development of NPQ. Under extreme conditions of excessive energy, sun leaves showed higher maximum DPS than shade leaves. Therefore, sun leaves contained not only a greater pool of xanthophyll cycle components but also a higher proportion of violaxanthin (V) actually photoconvertible to A and Z, compared to shade leaves. Both these effects contributed to the higher NPQ in sun versus shade leaves. The amount of photoconvertible V was strongly related to chla/b ratio and inversely to leaf neoxanthin content. This evidence indicates that the amount of photoconvertible V may be dependent on the degree of thylakoid membrane appression and on the organization of chlorophyll-protein complexes, and possible explanations are discussed. Exposure to chilling temperatures caused a strong decline in the photon yield of photosynthesis and in the intrinsic efficiency of PS II photochemistry in sun leaves, but little effects in shade leaves. These effects were accompanied by increases in the pool of xanthophyll cycle components and in DPS, more pronounced in sun than in shade leaves. This corroborates the view that Z and A may play a photoprotective role under unfavorable conditions. In addition to the xanthophyll-related non-radiative energy dissipation, a slow relaxing component of NPQ, independent from A and Z concentrations, has been found in leaves exposed to low temperature and high light. This quenching component may be attributed either to other regulatory mechanism of PS II efficiency or to photoinactivation.