Microorganisms represent the largest component of biodiversity in our biosphere. Traditional methods of bacterial identification depend on their culture on laboratory media and the comparison of their phenotypic characteristics. They include cellular morphology, motility, staining reactions of cell walls, ability to grow on different media and biochemical tests. These methods have many limitations and only a very small fraction of microorganisms have been cultivated. To date, molecular methods based on 16S rRNA sequences and their phylogenetic analysis are widely used for reliable identification, particularly for hard-to-culture microbial pathogens. These so-called << molecular methods >> do not require laboratory culture of isolated organisms, and many novel non-described phyla have been detected, improving our view of bacterial diversity. Novel strategies for culturing the << uncultivated >> are now under development, which are leading to the complete characterization of these new bacteria. More recently, meta- or ecogenomics, based on the complete sequencing of clones containing cosmids or bacterial artificial chromosomes with inserts, addresses the genetic potential of a sample irrespective of whether the microorganisms can be cultured or not. This has considerably extended our view of microbial diversity at the genomic level and the probability of finding new genes and their products suitable for the biotechnological and pharmaceutical industry.