The concept of induction followed by a long-term maintenance treatment has attracted much attention for the treatment of multiple sclerosis over the 30 past years. It was first demonstrated by the combination of induction therapy with mitoxantrone (six-monthly courses) followed by maintenance therapy with an immunomodulatory treatment such as an interferon-β or glatiramer acetate. Long-term observational studies confirmed that this therapeutic regimen provides a rapid reduction in disease activity and sustained disease control up to at least five years in 60% of patients. A better treatment response was observed in patients with early signs of aggressive disease, as shown in randomised studies (using six-monthly 12mg/m² of mitoxantrone intravenously at a cumulative dose of 72mg/m², followed by an interferon-β) as well as in long-term observational studies. But the safety profile of mitoxantrone make it more particularly suitable for young patients with frequent early relapses with incomplete recovery and multiple gadolinium-enhancing T1 lesions or spinal cord lesions on magnetic resonance imaging. More recently approved, the second candidate for an induction strategy is alemtuzumab: phases II and III randomised studies showed the superiority of alemtuzumab 12mg per day given intravenously for only five days and repeated for 3 days one year later, compared with interferon-β three times a week. Like with mitoxantrone, results supported the concept of long-term benefit after a short induction rather than escalation, in a subset of patients with early very active MS, with a sustained control of the disease for up to 7 years in 60% of patients in the phase III extension studies and in a long-term observational study. On the contrary, when alemtuzumab was first studied later in the disease course, results were disappointing. However, the risk of developing manageable but potentially severe systemic autoimmune diseases within the years following the last course of alemtuzumab make it, like mitoxantrone, more suitable for patients with early aggressive MS. More recently, cladribine an oral immunosuppressant, showed interesting results in a phase III study extension suggesting its potential induction effect, since after two cycles of treatment (5 days repeated 1 month later) at one year of interval, the remained low up to 4 years of follow-up, in the absence of any new treatment. However, today other immunosuppressive drugs have proved to be strongly and rapidly efficacious in treating highly active MS patients but through a mechanism of continuous immunosuppression (i.e., natalizumab and ocrelizumab). Indeed, disease activity can reappear rapidly after stopping these drugs, sometimes associated with a rebound of the inflammatory process, which is the contrary of a mechanism of induction that is associated with a remnant effect. Taking into account advantages and disadvantages of the different DMDs, which enriched the today therapeutic arsenal for MS, we propose in this paper some algorithms summarizing our reflexion about using an escalation strategy or an induction strategy according to disease course and activity.