Adaptation of rumen fermentation to monensin feeding has been studied with rumen-fistulated sheep receiving a daily dose of 30 mg of monensin for a period of 21 d followed by a 28 d period during which 60 mg doses were administered. The ration consisted of 300 g of hay and 300 g of concentrates, fed at 9.00 h and 16.00 h. Monensin was placed in the rumen as an aqueous suspension, just prior to the morning feeding. Monensin infusion was preceded and followed by a period during which no monensin was infused. The following rumen fermentation parameters were determined: methane production, pH, volatile fatty acids (VFA) molar proportions, total volatile fatty acid concentration, lactate and ammonia concentrations and in sacco degradability of hay. Rumen gas expelled through the fistula was collected for 6 h per day and analysed. Total VFA concentration, molar proportions of individual VFA, pH, lactate and ammonia concentration were determined on rumen contents, sampled just prior the administration of monensin and 2 and 6 h later. In vitro incubations of 3 h were carried out with rumen fluid, sampled 1 h after feeding. In vivo and in vitro methane production was decreased by monensin feeding. The molar proportion of propionate in the rumen was increased, while acetate and butyrate percentages were lowered. The total VFA and ammonia concentrations were also decreased by monensin, but pH values were increased. In vitro production of propionate was stimulated by monensin administration and methanogenesis decreased. The organic matter in sacco degradability was not affected, probably because of the time difference between the introduction of bags and monensin in the rumen. These modifications of rumen fermentation persisted as long as monensin was given, indicating that in this experiment, there was no adaptation to the ionophore.