Abstract It is generally accepted that behavioral interventions must follow systematic hypotheses regarding variables that maintain problem behavior. Hypotheses-based interventions are more likely to address behavioral functions and decrease or eliminate all problem behavior, often by teaching functionally equivalent appropriate responses as replacement behaviors. However, in some cases initial functional assessment/analyses may not lead to hypotheses that result in effective interventions. In such cases, it is important to continue functional analysis procedures to look more precisely at behavioral functions. This case study discusses how behavioral interventions were modified for a 27 year-old woman with severe disabilities, following systematic analyses of behavioral functions when initial interventions were ineffective. Keywords: Functional assessment, hypothesis-based intervention, response sequences, covariation ********** A thorough understanding of the relationships between behavior and ecological/environmental variables has long been the foundation for effective treatment of problem behavior in community settings (Kern, Childs, Dunlap, Clarke, & Falk, 1994; Mace & Lalli, 1991; Selinske, Greer, & Lodhi, 1991). This process has assumed even increased prominence with the current focus on further refinement of the functional assessment technology to ensure effective delivery of intervention (Borrero & Vollmer, 2002; Galiatsatos & Graff, 2003; Kennedy, Meyer, Knowles, & Shukla, 2000). Review of the professional literature makes it quite clear that best practice regarding the delivery of behavioral support requires the use of functional assessment and analysis procedures to guide the design of functionally appropriate intervention strategies (Carr, Robinson, & Palumbo, 1990; Day, Horner, & O'Neill, 1994; Dunlap, Kern-Dunlap, Clarke, & Robbins, 1991; Iwata, Vollmer, & Zarcone, 1990; Repp & Karsh, 1994; Sprague & Horner, 1992; Steege, Wacker, Berg, Cigrand, & Cooper, 1989). Recent reviews of the professional literature show that there has been some lag between identifying functional assessment as a best and essential practice in behavioral support, and the actual use of these procedures in reports appearing in the published literature (Blakslee, Sugai, & Gruba, 1994). Although this may reflect publication lag and may be changing (Blakslee et al., 1994), contacts with community practitioners, teachers, family members, and paraprofessionals at local and regional workshops and conferences suggest that a wide gap still exists between research and clinical practice in community settings. It remains very important that the factors that contribute to lack of use of functional assessment and support strategies may be identified. One such factor may be the difficulty encountered in some cases in identifying behavioral functions and generating adequate hypotheses to guide development of effective support interventions (Holden, 2002). Initial assessment information and initial hypotheses may not be accurate or may lack sufficient precision to guide effective intervention. Factors that may complicate functional assessment and hypothesis development include the fact that problem behaviors may serve multiple functions (Day, Horner, & O'Neill, 1994; Iwata, Pace, Cowdery, & Miltenberger, 1994), that problem behaviors may be controlled by contextual variables, some of which may not be easily identified or isolated (Carr, 1994; Haring K Smith, Iwata, Goh, & Shore, 1995), escalation in problem behavior may place the participants/therapist at risk (Smith & Churchill, 2002), that multiple topographies of problem behaviors within a response class may covary systematically, with the occurrence of some behaviors influencing the probability of occurrence of other, subsequent behaviors (Albin, O'Brien, & Horner, 1995; Kennedy, et al. …