To highly conserved primer binding sites, 16S rRNA gene sequences contain hypervariable regions that can provide species-specific signature sequences useful for identification of bacteria. As a result, 16S rRNA gene sequencing has become prevalent in medical microbiology as a rapid and cheap alternative to phenotypic methods of bacterial identification.

You can use databases like:                                                                                          1) EzTaxon-e. http://eztaxon-e.ezbiocloud.net/ The EzTaxon-e-database is an extension of the original EzTaxon database. It contains comprehensive 16S rRNA gene sequences of taxa with valid names as well as sequences of uncultured taxa. EzTaxon-e contains complete hierarchical taxonomic structure (from phylum rank to species rank) for the domain of bacteria and archaea.

2) Ribosomal Database Project. http://rdp.cme.msu.edu/ The Ribosomal Database Project (RDP) is a curated database that offers ribosome data along with related programs and services. 3) SILVA. SILVA provides comprehensive, quality checked and regularly updated datasets of aligned small (16S/18S, SSU) and large subunit (23S/28S, LSU) ribosomal RNA (rRNA) sequences for all three domains of life as well as a suite of search, primer-design and alignment tools (Bacteria,

3) SILVA. SILVA provides comprehensive, quality checked and regularly updated datasets of aligned small (16S/18S, SSU) and large subunit (23S/28S, LSU) ribosomal RNA (rRNA) sequences for all three domains of life as well as a suite of search, primer-design and alignment tools (Bacteria, Archaea and Eukarya).

4) Greengenes. Greengenes is a quality controlled, comprehensive 16S reference database and taxonomy based on a de novo phylogeny that provides standard operational taxonomic unit sets. 

In addition, you can take a look on this paper link; http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3592464/  The results of this study may be used as a guideline for selecting primer pairs with the best overall coverage and phylum spectrum for specific applications, therefore reducing the bias in PCR-based microbial diversity studies.

I second the suggestion of reading Klindworth et al. 

Since 16s databases are skewed toward microbes of interest to humans, members of the human microbiome should be well-represented (unlike, say, the microbiome of the deep ocean).  I have used both silva and RDP in the past; RDP allowed me to see if my primers missed particular taxa and thus required redesign.

I would use multiple primer sets. That's always a good approach.

As stated above read the Klindworth paper and also see what was used in the HMP http://hmpdacc.org/HM16STR/ if you want to compare data