This study aims to determine whether sheep modify their feeding and general behaviour when they undergo acidosis challenge, whether these modifications are maintained when acidosis challenges are repeated and whether yeast supplementation affects these modifications. Twelve rumen-cannulated wethers fed concentrate (wheat) and forage (hay) were exposed to three 28-day periods consisting of a 23-day recovery phase (20% of wheat) followed by a 5-day acidosis challenge (60% of wheat). Both diets limited food intake to 90% of ad libitum intake. Six sheep received a daily supplementation of a live yeast product, six received a placebo. Ruminal pH was recorded continuously. Daily consumption of wheat, hay, water and weekly consumption of salt were monitored. Behavioural observations were performed twice in each period: once under the recovery phase and once under acidosis challenge. These observations included video recordings over 24 h (time budget), social tests (mixing with another sheep for 5 min) and nociception tests (CO2 hot laser). As expected, sheep spent more time with a ruminal pH below 5.6 during challenges than during recovery phases (12.5 v. 4.7 h/day). Sheep drank more water (3.87 v. 3.27 l/day) and ingested more salt (16 v. 11 g/day) during challenges. They also spent more time standing than during recovery phases, adopting more frequent alarm postures and reacting more slowly to the hot stimulus. More severe behavioural modifications were observed during the first challenge than the two other challenges. Significant concentrate refusals were observed during challenge 1: from days 3 to 5 of this challenge, sheep ate only half of the distributed concentrate. Sheep were also more active and more aggressive towards each other in challenge 1. These behavioural modifications disappeared as the challenges were repeated: no behavioural modifications were observed between challenges and recovery phases during periods 2 and 3, and furthermore, sheep rapidly ate all the concentrate distributed during the third challenge. Focusing on the effects of yeast, the only differences registered between the two groups concerned ruminal pH, that is, mean ruminal pH values in the supplemented group were lower during the first challenge (5.11 v. 5.60) but higher during the third challenge (5.84 v. 5.28). In conclusion, our experiment suggests sheep can adapt to acidosis challenges, especially with yeast supplementation. Otherwise, ruminal pH values remained low during challenges, indicating that the modifications of general and feeding behaviour in subacute ruminal acidosis situations are not due exclusively to low ruminal pH values.