The representation of bacterial and archaeal genome sequences is strongly biased towards cultivated organisms, which belong to merely four phylogenetic groups. Functional information and inter-phylum level relationships are still largely underexplored for candidate phyla, which are often referred to as microbial dark matter. Furthermore, a large portion of the 16S rRNA gene records in the GenBank database are labeled as environmental samples and unclassified, which is in part due to low read accuracy, potential chimeric sequences produced during PCR amplifications and the low resolution of short amplicons. In order to improve the phylogenetic classification of novel species and advance our knowledge of the ecosystem function of uncultivated microorganisms, high-throughput full length 16S rRNA gene sequencing methodologies with reduced biases are needed. We evaluated the performance of PacBio single-molecule real-time (SMRT) sequencing in high-resolution phylogenetic microbial community profiling. For this purpose, we compared PacBio and Illumina metagenomic shotgun and 16S rRNA gene sequencing of a mock community as well as of an environmental sample from Sakinaw Lake, British Columbia. Sakinaw Lake is known to contain a large age of microbial species from candidate phyla. Sequencing results show that community structure based on PacBio shotgun and 16S rRNA gene sequences is highly similar in both the mock and the environmental communities. Resolution power and community representation accuracy from SMRT sequencing data appeared to be independent of GC content of microbial genomes and was higher when compared to Illumina-based metagenome shotgun and 16S rRNA gene (iTag) sequences, e.g. full-length sequencing resolved all 23 OTUs in the mock community, while iTags did not resolve closely related species. SMRT sequencing hence offers various potential benefits when characterizing uncharted microbial communities.