Abstract Background Obtaining accurate estimates of microbial diversity using rDNA profiling is the first step in most metagenomics projects. Consequently, most metagenomic projects spend considerable amounts of time, money and manpower for experimentally cloning, amplifying and sequencing the rDNA content in a metagenomic sample. In the second step, the entire genomic content of the metagenome is extracted, sequenced and analyzed. Since DNA sequences obtained in this second step also contain rDNA fragments, rapid in silico identification of these rDNA fragments would drastically reduce the cost, time and effort of current metagenomic projects by entirely bypassing the experimental steps of primer based rDNA amplification, cloning and sequencing. In this study, we present an algorithm called i-rDNA that can facilitate the rapid detection of 16S rDNA fragments from amongst millions of sequences in metagenomic data sets with high detection sensitivity. Results Performance evaluation with data sets/database variants simulating typical metagenomic scenarios indicates the significantly high detection sensitivity of i-rDNA. Moreover, i-rDNA can process a million sequences in less than an hour on a simple desktop with modest hardware specifications. Conclusions In addition to the speed of execution, high sensitivity and low false positive rate, the utility of the algorithmic approach discussed in this paper is immense given that it would help in bypassing the entire experimental step of primer-based rDNA amplification, cloning and sequencing. Application of this algorithmic approach would thus drastically reduce the cost, time and human efforts invested in all metagenomic projects. Availability A web-server for the i-rDNA algorithm is available at http://metagenomics.atc.tcs.com/i-rDNA/