OBJECTIVES: To determine the long-range (4 to 10 years) cost burden of several possible treatment sequences of all drugs recommended in Portugal for the treatment of highly active relapsing multiple sclerosis, using as effectiveness measure the time to first relapse published in clinical or observational long-term studies. METHODS: An Excel simulation model was developed to determine the time on treatment and the incurred costs of individual patients treated with MAVENCLAD^®, fingolimod and natalizumab. Time on treatment was randomly picked from the time to first relapse survival curves available in literature for the three products and impact was evaluated at 4, 6, 8 and 10 years horizons. Patients stayed on each treatment for no longer than 4 years, after which ther rotated to the next therapy. If total treatment with the three drugs was less than the horizon, the remaining period was assumed as treated with interferon b1a 44mg. Costs were obtained from official Portuguese data sources. RESULTS: A run of 10.000 Monte Carlo simulations of high disease activity patients, showed that the lowest budget impact was obtained with the sequence MAVENCLAD^® -natalizumab-fingolimod for 4 to 10 year time horizons (60.438,37 €, 93.831,59 €, 126.757,90 € and 158.436,19 €, respectively), whereas highest burden was found for fingolimod-natalizumab-MAVENCLAD^® at 4 and 6 year horizons (98.539,61 € and 128.494,33 €, respectively) and for natalizumab-fingolimod-MAVENCLAD^® at 8 and 10 year horizons (152.127,72 € and 172.575,13 €, respectively). CONCLUSIONS: In the current model starting with MAVENCLAD^® seems to offer better budget impacts over 10 years than starting treatment with either fingolimod or natalizumab. These results should be taken into account when informing medical decision-making.