Mutants of the unicellular, green alga Chlamydomonas reinhardtii were recovered by screening for the absence of photoautotrophic growth at 35°C. Whereas nonconditional mutants required acetate for growth at both 25 and 35°C, the conditional mutants have normal photoautotrophic growth at 25°C. The conditional mutants consisted of two classes: (a) Temperature-sensitive mutants died under all growth conditions at 35°C, but (b) temperature-sensitive, acetate-requiring mutants were capable of heterotrophic growth at 35°C when supplied with acetate in the dark. The majority of mutants within the latter of these two classes had defects in photosynthetic functions. These defects included altered pigmentation, reduced whole-chain electron-transport activity, reduced ribulosebisphosphate carboxylase activity, or pleiotropic alterations in a number of these photosynthetic components. Both nuclear and chloroplast mutants were identified, and a correlation between light-sensitive and photosynthesis-deficient phenotypes was observed.The unicellular green alga Chlanmydomtonas reinhardtii has proven to be a useful organism for the genetic dissection and analysis of photosynthesis in eukaryotes (16). Since C. reinhardtii maintains a functional photosynthetic apparatus when grown in the dark with acetate, photosynthesis-deficient mutants have been identified as light-sensitive, acetate-requiring strains (10). Genetic analysis can also be performed, and mendelian or uniparental (maternal) inheritance defines nuclear or chloroplast mutations, respectively (5).We have been particularly interested in using chloroplast mutants and their revertants to investigate the structure/function relationships of Rubisco& (13). As in higher plants, Rubisco consists of eight nuclear-encoded small subunits and eight chloroplast-encoded large subunits, and the large subunit appears to contain the active-site domain (7). However, of three Rubisco mutants defined at the level of the large-subunit gene sequence (3, 15), two were found to result from nonsense mutations. Calculating from the large-subunit gene sequence (2), one would expect only about 5% nonsense mutations out of the total number of point mutations that could alter the large-subunit protein sequence. It thus appeared that many mutants with amino-acid substitutions in the large-subunit protein were not being re-