Can J Psychiatry. 2011;56(11):641-642. Since the publication of the seminal text 2 decades ago,' motivational interviewing (Ml) has seen a steady rise in popularity in the behavioural health community. In 2 decades, it has grown from a treatment that was focused solely on unmotivated substance abusers to a mainstream behavioural intervention adapted for various health-related and psychiatric problems. Information from dissemination theory2 indicates that MI is relatively early in the adoption curve, so we can expect that it will continue to be a vital area of research and clinical practice for some time to come. The rising tide of MI dissemination and adaptation proceeds, in many ways, without critical science to allow its thoughtful and prudent use. For example, there is no answer in sight concerning the question of what problem is not amenable to Ml. IfMI is indeed an efficacious behavioural intervention (as opposed to a panacea or placebo) it must surely have some boundaries to its effectiveness, yet what are they? The current state of the research base shows that this treatment works moderately well for almost every problem area in which it is applied, and also that there is often no difference between MI and other treatments.3 How then does the well-intentioned clinician make plans to use, or set aside, this method? The 2 In Review articles4·5 in this issue will help the reader to answer this question by examining both the available research within a specific area (adolescents and psychiatric disorders) and the advantages and limitations inherent in applying MI outside its original domain of substance abuse. One theme that recurs when MI grows away from the substance abuse problems that were its home is the question of whether client ambivalence is a necessary condition for using MI, and if so, how much ambivalence must be present to begin. Formal MI theory assigns client ambivalence a central role in facilitating motivation for change.6 Interviewers are seen as eliciting rather than creating motivation, which instead arises from the resolution of ambivalence. But what happens when the client is genuinely unconcerned about changing and seeks treatment only to avoid aversive consequences? What if clients have very little ambivalence about self-destructive behaviours? How far can a treatment that was developed within a collaborative, humanistic psychotherapeutic context be adapted to situations where the client views him or herself as being manipulated into making a change that is not one they would ever likely choose? This issue is particularly relevant to psychiatric populations where ambivalence may represent a core characteristic of the disorder, such as schizophrenia or borderline personality disorder. It is reasonable to ask what results we might expect when selecting a treatment with the explicit goal of enhancing and enlarging ambivalence for a population in whom ambivalence is a chronic and enduring feature of their inner life. Similarly, how do we expect that MI will work with adolescents who often show very little ambivalence about the problems that bring them to the attention of concerned others? The In Review articles4·5 deal directly with this issue by presenting the research showing that MI is indeed effective, both for clients with severe mental illness and for adolescents. Both articles4·5 offer hints about the ways that MI must be tailored for these groups, and readers will find especially helpful the hands-on and practical examples, including dialogue, of how to approach adolescents using an MI framework. A second theme in the current Zeitgeist of MI is how to integrate it with other treatments. MI is now commonly married with cognitive-behavioural therapy (CBT) for various problems, and indeed almost never occurs as a stand-alone intervention outside the substance abuse treatment arena. The question of how MI can be combined with CBT necessarily requires us to consider how MI works, and whether it may carry intentions that are not compatible with other approaches. …