The effect of acclimation to 25, 18, or 10-degrees-C on the relationship between photoprotection and photodamage was tested in low-light-grown (80 mumol m-2.s-1) Ulva rotundata Blid. exposed to several higher irradiances at the acclimation temperature. Changes in chlorophyll fluorescence parameters (minimum fluorescence, F0, and the ratio of variable to maximum fluorescence, F(v)/F(m), measured after 5 min darkness) were monitored during 5 h transfers to 350, 850, and 1700 mumol. m-2 . S-1, and during recovery after 1- or 5-h treatments. At all temperatures, rate of onset and final extent of photoinhibition, measured by a decrease in F(v)/F(m), increased with increasing irradiance. At a given photoinhibitory irradiance, rate of onset was most rapid at 10-degrees-C, but the extent was temperature-independent. Recovery rates from mild light stress were similar at all temperatures, but recovery from the most extreme photoinhibitory treatment lagged 2 h at 10-degrees-C. De-epoxidation of xanthophyll-cycle components proceeded faster and to a lower epoxidation status at 25-degrees-C, but there was little difference in the pool size among the three growth conditions. Using chloramphenicol to inhibit chloroplast protein synthesis and dithiothreitol to inhibit violaxanthin de-epoxidation, it was shown that at the lowest light treatment given, the extent of photoinhibition could be attributed both to greater amounts of photodamage and to greater zeaxanthin-related photoprotection at 25 than at 10-degrees-C. While these two mechanisms for high-light-induced loss of photosynthetic efficiency were operating at 10-degrees-C, there was evidence for a relatively greater proportion of zeaxanthin-unrelated photoprotection at the low temperature. This photoprotective mechanism is related to a rapidly reversible increase in F0 and is insentivite to both chloramphenicol and dithiothreitol.