Pyrosequencing of plastid 23S rRNA genes was performed to determine the usefulness of this methodology for describing spatial and temporal patterns of algal diversity in two eutrophic lakes. The majority of the sequences were identified as known cyanobacteria or eukaryotic algae (> 70% of sequence reads), indicating this approach can specifically recover algal sequences from complex communities. Furthermore, estimated coverage of the data sets indicated that the majority of the 23S rRNA genetic diversity was recovered in these surveys. Communities from algal mats could be clearly distinguished from algae in the water column, and the communities could be readily differentiated between the two lakes, suggesting that the plastid 23S rRNA sequencing was able to distinguish niche and biogeographic partitioning of algal communities. Within the sequence data sets, the ratio of cyanobacteria to eukaryotic algae fluctuated over the course of sampling, with cyanobacteria 23S rRNA sequences being more abundant in later samples. In addition, the eukaryotic algae communities showed large shifts in composition over the course of sampling. Taken together, these data demonstrate the usefulness of targeted plastid 23S rRNA sequencing for describing the structure and dynamics of complex algal communities.