An in vitro experiment was designed to investigate the effects of incubating two forages with a different energy/nitrogen (N) ratio [perennial ryegrass (GR) vs red clover (RC)] on the efficiency of N utilisation by rumen microbes. Second-cut forages were incubated in artificial rumen fermenters (n = 8). Ryegrass represented a supply of quickly available N and energy for the rumen microorganism which led to a rapid fermentation and bacterial growth 2–4 h after feeding. Ryegrass also promoted greater numbers of anaerobic fungi, methanogens and cellulolytic bacteria, which tended to increase neutral detergent fibre disappearance, gas production, volatile fatty acid and methane production than observed using RC diets. On the contrary, RC provided slowly degradable N and energy, which led to a relatively slow bacterial growth (4–8 h after feeding). In terms of diet utilisation, RC diets promoted a higher N outflow (mainly as undegraded-N) and efficiency of microbial protein synthesis per organic matter disappeared. Even so, microbial protein yield was similar on both diets indicating a better N capture by microorganisms fed GR than in those fed RC diets. The use of 15N-labelled forages demonstrated that this high ammonia incorporation by bacteria-fed GR occurred mainly during the early fermentation coinciding with the highest bacterial growth. In conclusion, this experiment demonstrated that the use of isotopic labelling combined with molecular techniques provided an insight into forage utilisation by the rumen microbes; GR diets led to a better efficiency of N utilisation compared with RC; moreover the lower N outflow on GR diets may be partially compensated for a higher proportion of microbial protein leaving the system and the greater volatile fatty acid production. These findings seem to indicate that RC grazing may increase the N pollution compared with GR without substantial improvements on the rumen function, however this must be confirmed in vivo.