Rapid and precise identification of microbial species from a mixture of microbes is the key to developing novel treatments for infectious diseases. Such techniques also provide valuable tools to the study of tissue or environmental microbiome with high diversity. Fluorescencein situhybridization (FISH) technique allows to detect and identify microbes based on the variation in their genome sequence without the need for time-consuming culturing or sequencing. However, this approach is restricted by crosstalk between species due to non-specific binding of FISH probes. We developed a novel set of FISH probes through large scale analysis of 16S sequence variation in microbiome and minimizing crosstalk between species. Using these probes, FISH measurements verified that 8 microbial species could be successfully distinguished. Based on the database, we subsequently developed peptide nucleic acid (PNA) FISH probes which allowed rapid and efficient hybridization. As PNA probes allowed to use shorter probe length, small sequence variation between close species could be maximally utilized to produce large contrast. Using these sets of probes, we further developed sequential FISH techniques that allowed precise identification of species by rejecting possible errors and crosstalk from a single set of probes..