We examined the clinical course after switching disease-modifying therapy (DMT) in patients with relapsing–remitting multiple sclerosis (RRMS). Eighty-five consecutive RRMS patients who received weekly interferon beta-1a (IFN beta-1a) 6 MU i.m. for at least 18 months were enrolled. Baseline annualized relapse rate (ARR) for the 2 years prior to initiating therapy with IFN beta-1a was obtained from charts. All 85 patients received treatment with IFN beta-1a at 6 MU i.m. weekly for 18–24 months (mean 19.7 months). Treatment with IFN beta-1a reduced the mean ARR from 1.41 to 1.23 (P = 0.005). All 85 patients were then switched to glatiramer acetate (GA) 20 mg s.c. daily and prospectively followed up for 36–42 months (mean 37.5 months). Patients were switched because of persistent clinical disease activity (n = 62) or persistently unacceptable toxicity (n = 23) as determined by the treating neurologist. Treatment with GA reduced the mean ARR from 1.23 to 0.53 (P = 0.0001). Subgroup analysis showed that in patients who were switched because of lack of efficacy (n = 62), the mean ARR was reduced from 1.32 on IFN beta-1a to 0.52 on GA (P = 0.0001). In contrast, in patients who switched because of persistent toxicity (n = 23), the mean ARR was reduced from 0.61 on IFN beta-1a to 0.47 on GA (P, non-significant). Our observations suggest that clinical observations such as relapse rate and tolerability may be used as criteria for switching DMT in clinical practice. More definitive consensus criteria incorporating magnetic resonance imaging and clinical observations for defining optimal response and tolerability need to be developed for the routine clinical management of RRMS patients receiving DMT.