The article by Stevenson and colleagues1 advances our understanding of the effectiveness of constraint-induced movement therapy (CIMT) as a technique to improve upper-limb (UL) recovery post stroke. Comparing CIMT to dose-matched control groups is vital in determining key therapeutic ingredients for clinicians. Time spent doing therapeutic activity has been studied and is regarded as an important element of stroke intervention; findings by Stevenson and colleagues suggest, however, that critical elements of therapy may go beyond time. The studies by Taub and colleagues2 and Wolf and colleagues3 that introduced CIMT for recovery of UL function post stroke were pivotal. The theory of non-use illustrated by primate studies,4 operant conditioning (shaping), and high-intensity protocols were new ideas for treating individuals with stroke. The focus was on forcing the use of the affected limb during mass practice of activities, including grading of activities to match motor capability; there was little focus on inhibiting movement or facilitating normal movement patterns. The client was an active participant in the programme, not a passive recipient of therapist-driven movement. The mounting evidence to support the effectiveness of CIMT compared to traditional treatment5 represented a paradigm shift in UL intervention strategies. The protocol of the original CIMT (6 h/d for 2 wk, less-affected UL constrained for 90% of waking time) was not deemed clinically feasible by clinicians and researchers. However, the strength of the findings prompted investigators to examine modified protocols more suitable for clinical settings. Results continued to support the superiority of CIMT over usual/standard care6 (often defined as weight-bearing activities on the affected side, facilitation of movement, and compensatory strategies using the less affected UL). Despite this evidence and resulting best practice standards,7 however, the implementation of CIMT in current practice is limited. The most prevalent barrier reported by therapists is time,8 in that even a 2-hour protocol is not feasible in an in-patient setting or in many outpatient settings. Stevenson and colleagues1 bring our attention to an important element of UL intervention strategies: dosage. Stevenson and colleagues' findings suggest that when dosage is equal, CIMT results in greater UL function and participation in activities of daily living compared to standard interventions.1 This is a critical finding. We are now able to further articulate key therapeutic ingredients for improved UL function post stroke. Whether the key factor is the activity (relevance, complexity), strategies (e.g., forced use, shaping), and/or environment (e.g., challenging, stimulating), the answer appears to go beyond the amount of time spent completing activities. The take-home message for clinicians is that strategies that enhance the intensity of the activity (e.g., repetitions, complexity, and effort), force the use of the affected limb, and foster its use outside of therapy time are key to effective treatment. Another interesting finding from Stevenson and colleagues is that stroke acuity did not appear to be a factor in improved outcomes for CIMT.1 Previous studies have suggested that CIMT is not more effective in the acute stages of stroke recovery,9 opening the door for more research with this group. It is important to note that CIMT is not appropriate for all people with stroke; in fact, it is estimated that <10% of people with stroke meet the inclusion criteria.10 For the majority of people admitted to rehabilitation units and outpatient clinics or receiving community-based care, current models of care may suffice. However, as Stevenson and colleagues1 conclude, the challenge remains of how to translate CIMT protocols into clinical practice. Barriers to implementation still exist, despite mounting evidence that CIMT should be considered a key intervention for UL recovery post stroke. This issue goes beyond the individual therapist and must also be considered a matter of administrative concern: if one of the most effective currently available means of improving UL recovery post stroke, and thus perhaps decreasing time in care and enhancing quality of life, is 2 hours/day of CIMT, 5 days/week, for 2 weeks, then therapists should be given the means to implement best practice. The way to do this is through collaboration among policy makers, administrators, clinicians, researchers, and—most critically—clients and family members.