AIMS: This study evaluated the effects of DNA extraction method, DNA purification and pooling of PCR amplification products on the description of bacterial and archaeal diversity. METHODS AND RESULTS: Soil DNA was extracted by the Power Soil DNA extraction kit and a customized Griffiths' protocol. Both methods are based on cell disruption by bead beating. In total, we used three soils and six independent extractions from each soil obtained by each of the two methods. Then, three of the six extracts of each treatment were further purified by spin columns filled with Sepharose 2B and polyvinylpolypyrrolidone (PVPP). The V4 hypervariable region of the 16S rRNA gene was amplified from each extract using the 515F/806R primer pair in four independent reactions. Three amplification products were combined and sequenced as a pooled sample, while the additional amplification product was sequenced individually. The resulting 72 amplification products were sequenced by Illumina MiSeq platform. DNA extraction method had a statistically significant effect on the estimation of the composition of microbial communities that might overwhelm differences in microbial communities from distinct soils. On the other hand, a further DNA purification step or pooling of PCR amplification products had a minor effect on the description of bacterial and archaeal communities. CONCLUSIONS: DNA extraction had the strongest effect on the description of bacterial and archaeal communities; low concentration of impurities, which allow PCR amplification, can still generate a minor additional bias, while PCR stochastic variability had the lowest effect. SIGNIFICANCE AND IMPACT OF THE STUDY: Although it is well known that methodological factors affect the description of microbial communities, the relative importance of each step is still unknown. The present study determined that of the factors tested, the DNA extraction method had the strongest effects on the description of bacterial and archaeal communities.