•Melbourne, Australia experienced prolonged, repeated and strict lockdowns.•Effects on exercise in MS varied according to lockdown severity, disability and technology use.•Exercise frequency dropped more in exercisers with mild than exercisers with moderate disability.•Conversely, sedentary behaviour decreased significantly more in extended, severe lockdowns.•Home exercise increased during lockdowns and was associated with use of on-line exercise classes. Regular exercise is beneficial for people with Multiple Sclerosis (MS), regardless of disability level. The previously reported differential effect of COVID-19-related lockdowns on exercise levels in this population remains unexplained. We examined effects of lockdowns on exercise in Australians with MS according to disability levels, lockdown severity and health technology use. A cross-sectional survey of people with MS in Australia (22 April-23 September 2021) collected demographic and clinical information as well as exercise patterns before and during lockdowns. Mann-Whitney was used to compare ordinal data and Likelihood Ratio to compare dichotomous data. 151 people completed the survey. 72.2% had mild disability and 25.2% moderate disability. Extended lockdowns were associated with significantly decreased sedentary behaviour (31.5% to 25.9%) but also with decreased exercise frequency in frequent exercisers (≥3 times/week; 53.7% to 22.2%). The latter occurred significantly more in those with mild disability (-22.7%) than with moderate disability (-3.5%). More people with mild disability walked for exercise pre-pandemic (LR 8.6, p=.004) and during lockdowns (LR 6.6, p=.010). Walking during lockdowns was positively associated with working from home. People with moderate disability were more likely to engage in home exercise both pre-pandemic (LR 5.5, p=.019) and during lockdown (LR 5.2, p=.023). Engagement in home exercise rose for both groups during lockdowns and was facilitated by on-line exercise classes. Lockdowns differentially affected exercise patterns according to disability level. The proportion of people achieving exercise recommendations decreased more in those with mild but not moderate disability. Incidental physical activity was disproportionately impacted in people with moderate disability.