The objective was to evaluate the effects of an inhibitor of α-amylase and glucosidase (acarbose, Pfizer Limited, Corby, UK) on ruminal fermentation, blood metabolism and microbial profile in dairy cows in a 2×2 cross-over experiment. Eight Holstein cows fitted with rumen cannulas (milk yield, 24·3±2·35 kg/d, body weight, 622±54 kg, days in milk, 183±67, 5 multiparous and 3 primiparous) were used. Treatments were: control (no additive, CTR) and α-amylase and glucosidase inhibitor (0·75 g acarbose-premix/cow per d, AMI). Animals were given ad-libitum access to a high non-fibre carbohydrate (NFC) partial mixed ration (PMR) containing 17·6% crude protein, 28·3% neutral detergent fibre, and 46·5% NFC in the dry matter and supplementary concentrate during milking. Blood samples were taken to determine blood glucose, insulin and urea within the first hour after the morning feeding on two separate days in each period. Samples of ruminal contents were collected during 3 d in each period at 0, 4 and 8 h after feeding to determine volatile fatty acid and ammonia-N concentrations and to quantify protozoa, Streptococcus bovis and Megasphaera elsdenii. Rumen pH was recorded electronically at 22-min intervals during 6 d in each period. Results were analysed using a mixed-effects model. Cows on AMI treatment spent less time with ruminal pH <5·6 compared with cows in the CTR group (3·74 and 6·52±0·704 h/d, respectively). Cows in the AMI group had greater daily average pH compared with those in the CTR group (6·05 and 5·92±0·042, respectively). AMI animals tended (P=0·09) to have lower Str. bovis to Meg. elsdenii ratio than CTR (4·09 and 26·8±12·0, respectively). These results indicate that dietary supplementation with acarbose in dairy cattle fed high-production rations may be effective in reducing the time for which rumen pH is suboptimal, with no negative effects on ruminal fermentation and blood metabolites.