Understanding, predicting, and changing behavior is critical to all facets of life, especially when behavior is considered problematic to the individual behaving or to those around him/her. Our efforts to resolve behavior problems are greatly hindered when behavior is conceptualized as something an individual has rather than something s/he does under certain conditions and not under other conditions. When we think that problem behaviors are due to something inside the individual, we naturally consider it the in-dividual’s problem. When we think that problem behaviors are due to the conditions in which the behavior is demonstrated, we try to change the conditions — that part of behavior we can do something about. In this presentation, a cross-species model for assessing how conditions set the occasion for and maintain prob-lem behaviors will be discussed. This model, called functional as-sessment, reveals answers to the three fundamental behavior-change questions: What (identify the problem behavior in ob-servable, unambiguous terms); when (predict the conditions un-der which it will occur and not occur); and, why (what purpose does it serve for the individual). The far-reaching relevance of this model to improving the lives of all individuals, human and non-human, will be discussed.