Rusitec fermenters are in vitro systems widely used to study ruminal fermentation, but little is known about the microbial populations establishing in them. This study was designed to assess the time evolution of microbial populations in fermenters fed medium- (MC; 50% alfalfa hay : concentrate) and high-concentrate diets (HC; 15 : 85 barley straw : concentrate). Samples from solid (SOL) and liquid (LIQ) content of fermenters were taken immediately before feeding on days 3, 8 and 14 of incubation for quantitative polymerase chain reaction and automated ribosomal intergenic spacer analysis analyses. In SOL, total bacterial DNA concentration and relative abundance of Ruminococcus flavefaciens remained unchanged over the incubation period, but protozoal DNA concentration and abundance of Fibrobacter succinogenes, Ruminococcus albus and fungi decreased and abundance of methanogenic archaea increased. In LIQ, total bacterial DNA concentration increased with time, whereas concentration of protozoal DNA and abundance of methanogens and fungi decreased. Diet×time interactions were observed for bacterial and protozoal DNA and relative abundance of F. succinogenes and R. albus in SOL, as well as for protozoal DNA in LIQ. Bacterial diversity in SOL increased with time, but no changes were observed in LIQ. The incubated diet influenced all microbial populations, with the exception of total bacteria and fungi abundance in LIQ. Bacterial diversity was higher in MC-fed than in HC-fed fermenters in SOL, but no differences were detected in LIQ. Values of pH, daily production of volatile fatty acids and CH4 and isobutyrate proportions remained stable over the incubation period, but other fermentation parameters varied with time. The relationships among microbial populations and fermentation parameters were in well agreement with those previously reported in in vivo studies. Using 15N as a microbial marker or quantifying total microbial DNA for estimating microbial protein synthesis offered similar results for diets comparison, but both methods presented contrasting results for microbial growth in SOL and LIQ phases. The study showed that fermentation parameters remained fairly stable over the commonly used sampling period (days 8 to 14), but shifts in microbial populations were detected. Moreover, microbial populations differed markedly from those in the inocula, which indicates the difficulty of directly transposing results on microbial populations developed in Rusitec fermenters to in vivo conditions.