The aim was to show the application of the prevalence estimator of Expanded Disability Status Scale (EDSS) improvement over time in patients treated with cladribine tablets in the phase III CLARITY/CLARITY extension trials. Relapsing-remitting multiple sclerosis patients who entered the CLARITY extension study were evaluated. Patients originally randomized in CLARITY to cladribine tablets 3.5 mg/kg and placebo in CLARITY extension (early cladribine [EC]) were compared to patients originally randomized to placebo and then assigned to cladribine tablets 3.5 mg/kg (delayed cladribine [DC]). The EC group was compared to the DC group on the prevalence of EDSS improvement over time and on the cumulative incidence of EDSS improvement. Prevalence of improvement was assessed by a new approach based on the difference of Kaplan-Meier estimators, whilst the incidence of improvement was assessed by standard Kaplan-Meier curves. A total of 98 patients in the EC group and 244 patients in the DC group were compared. Patients in the EC group showed a significantly higher (p = 0.011) prevalence of improvement at year 2 (EC 21.3%, 95% confidence interval [CI] 13.6-29.3; DC 8.9%, 95% CI 5.5-12.8) and at year 5 (EC 15.7%, 95% CI 8.2-23.7; DC 8.3%, 95% CI 4.5-12.4). The cumulative incidence of improvement was also significantly different (hazard ratio 1.82, 95% CI 1.13-2.94, p = 0.013). Assessment of the prevalence of EDSS improvement is an alternative outcome to assess if a treatment induces and maintains an improvement over the long term. This estimator was found to be more powerful than the cumulative incidence of improvement to detect a treatment effect of cladribine versus placebo over 5 years.