OBJECTIVE : Treatment for progressive multiple sclerosis (pMS) is a key area of research. To date, whether MS type and the rehabilitation setting are associated with worse or better response to rehabilitation is unclear. We aimed to understand the association between balance and MS type, in/outpatient treatment and specificity of the intervention. METHODS : We assessed 150 people with MS before and after in/outpatient rehabilitation. The Berg Balance Scale (BBS) was used to discriminate between responders (≥+3-point improvement in BBS score; a clinically meaningful improvement) and non-responders to specific or non-specific balance rehabilitation. Factors associated with balance were analyzed by univariate and multivariable logistic regression analyses, estimating odds ratios (ORs) and 95% confidence intervals (CIs). RESULTS : Balance improved after rehabilitation: median (quartile 1 [Q1]-Q3) BBS score pre- and post-rehabilitation of 49 (45-53) and 52 (47-55) (P<0.001). Univariate logistic analysis revealed a clinically meaningful improvement in balance associated with pMS (OR 2.21 [95% CI 1.09-4.05]), inpatient therapy (0.41 [0.19-0.84]), using a walking aid (1.68 [1.06-2.69]), and low baseline BBS score (0.86 [0.81-0.92]). On multivariable analysis, probability of improvement was similar for participants with pMS and the relapsing-remitting form but was associated with low baseline BBS score and specific treatment (OR 0.81 [95% CI 0.74-0.89] and 5.66 [1.79-21.5]). CONCLUSIONS : A clinically meaningful improvement in balance was more likely when MS individuals with moderate to high disability had specific exercises targeting balance, but MS type did not influence the outcome.