BACKGROUND: Evidence suggests that exercise can alleviate symptoms associated with multiple sclerosis (MS). However, it is unclear whether exercise alleviates symptoms of anxiety, present in one-third of people with MS. This systematic review and meta-analysis investigates whether exercise interventions are effective in reducing anxiety in people with MS. METHODS: CINAHL, Embase, MEDLINE, PsycINFO, and the Cochrane Controlled Clinical Trials Register were searched to identify relevant randomized controlled trials (RCTs) published up to March 30, 2018. The primary outcome was postintervention difference in anxiety scores between intervention and control groups. Using a restricted maximum likelihood random-effects model, standardized mean differences were pooled and heterogeneity was estimated. Risk of bias was assessed using an a priori modified Cochrane Risk of Bias 2.0 tool. RESULTS: Four RCTs with 133 participants were included. Exercise type included walking, aerobic activities, and cycling. Intervention duration ranged from 8 to 26 weeks. The pooled standardized mean difference in anxiety score was -0.16 (95% CI, -0.50 to 0.19), without heterogeneity. Two studies had high risk of bias, and two had some risk of bias. CONCLUSIONS: Based on available RCT data, exercise does not seem to have an effect on anxiety in people with MS. These results should be interpreted with caution because studies had concerns of bias and small sample sizes, and anxiety was not the primary outcome measure. Future studies should exhibit sufficient reporting standards, alongside publishing protocols. There is opportunity to investigate the effect of exercise in people with MS experiencing clinically relevant levels of anxiety rather than in the general MS population.