Background: The development of evidence-based practice, which is increasingly popular in stuttering treatment, is closely linked to the development of outcome measures. Aims: Two approaches to the development of stuttering treatment outcome measures are outlined. The first is the deductive, top-down approach, where the development of specific outcome measures is guided by a priori general conceptualizations of the nature of the disorder. A competing approach to the development of stuttering treatment outcome measures is outlined. This is the inductive, or bottom up approach. This approach uses Baer's (1988, 1990) notion of specific complaints of clients as the starting point to develop inductive statements for use as general guidance for developing treatment outcome measures. Main Contribution: It is argued that the deductive approach to development of outcome measures has limitations. It is overly prescriptive, generating numerous and increasingly complex outcome measures that are potentially confusing for clinicians. Further, it is arbitrary and fragile, being linked to conceptualizations and theories about stuttering, which, by necessity, are limited themselves. Further, the development of numerous outcome measures is not compatible with the conduct of randomized controlled trials, which allow a maximum of two primary outcome measures. In contrast, the inductive approach to the development of outcome measures has in its favour that it is empirically driven rather than arbitrary, and that it facilitates a yoking of the development of clinical outcomes and the clinical methods to attain those outcomes. The approach is unlikely to lead to the development of fruitless treatment methods. Further, the approach is parsimonious to the extent that it is likely to produce few guiding generalities for treatment outcome assessment--perhaps as few as two in the case of adults and one in the case of preschoolers. This is well suited to the use of the randomized controlled trial as a source of evidence for treatment efficacy. Conclusions: One inductive statement that can be used to guide the development of outcome measures is that the ill effects of stuttered speech could be troubling for those who seek clinical help. The other is that those who seek clinical help are likely to experience speech-related anxiety. Together, these sources of information provide sound guidance for the development of outcome measures relating to stuttered speech and speech-related anxiety, and guidance for the development of treatments to offset those ill effects of the condition. Until another outcome measure can be derived from inductive processes, those treatment developments should serve all the needs of those who stutter and the clinicians who provide those needs. (Contains 1 note.)