ABSTRACT: Reports of widespread proteorhodopsin and bacteriochlorophyll a genes suggest that light may facilitate the growth of many marine bacteria. We tested this hypothesis by conducting 3 light manipulation experiments with bacterial communities obtained off the Southern California coast. In each experiment, duplicate 20 l mesocosms were subjected to either 14:10 h light:dark or continual darkness for 5 to 10 d. Automated Ribosomal Intergenic Spacer Analysis (ARISA), a whole-communityfingerprinting technique, was used to determine how light affects bacterial community structure. Light removal resulted in only minor changes in ARISA profiles (mean decrease in similarity between treatments: 6 ± 1% via Sorenson’s index [only considers presence/absence of taxa], 15 ± 7% via Pearson’s coefficient [considers relative abundance of taxa]) at the conclusion of the experiments. Oligotrophic communities responded nearly twice as strongly to light removal compared to mesotrophic communities. Phototrophs such as cyanobacteria exhibited consistent, sharp declines in dark treatments, indicating whatever heterotrophic abilities they possess were not sufficient to sustain them when faced with natural removal processes suchas grazing or viral lysis. Members of the broad SAR86, SAR11, SAR116, CFB and Roseobacter phylogenetic groups exhibited minor, mixed responses to light removal, suggesting that only a few select members may rely on phototrophy to a measurable extent. Our results indicate that the majority of non-cyanobacterial bacteria in the ocean do not depend heavily on light to maintain themselves over 5 or 10 d periods in the presence of natural removal processes, and hint that proteorhodopsin and bacteriochlorophyll a genes may have alternative, undetermined ecological benefits besides phototrophy.