Dissection of prokaryotic community structure is prerequisite to understand their ecological roles. Various methods are available for such a purpose which amplification and sequencing of 16S rRNA genes gained its popularity. However, conventional methods based on Sanger sequencing technique require cloning process prior to sequencing, and are expensive and labor-intensive. We investigated prokaryotic community structure in tidal flat sediments, Korea, using pyrosequencing and a subsequent automated bioinformatic pipeline for the rapid and accurate taxonomic assignment of each amplicon. The combination of pyrosequencing and bioinformatic analysis showed that bacterial and archaeal communities were more diverse than previously reported in clone library studies. Pyrosequencing analysis revealed 21 bacterial divisions and 37 candidate divisions. Proteobacteria was the most abundant division in the bacterial community, of which Gamma- and Delta-Proteobacteria were the most abundant. Similarly, 4 archaeal divisions were found in tidal flat sediments. Euryarchaeota was the most abundant division in the archaeal sequences, which was further divided into 8 classes and 11 unclassified euryarchaeota groups. The system developed here provides a simple, in-depth and automated way of dissecting a prokaryotic community structure without extensive pretreatment such as cloning.