Basic response–event relations can be difficult to demonstrate with individuals with profound multiple impairments who participate in adaptive switch programs. Numerous variables that impede learning and contribute to performance variability present challenges to assessment methods. As switch use may be the only operant under stimulus control with some individuals, however, improvement in assessment methods and measurement technology is important. In three case studies, response–event relations were examined in individuals who were provided with an adaptive switch that could activate an electronic leisure device. With two individuals, evidence of learning was assessed in a novel alternating-treatments design. Switch use was compared under conditions in which switch closures produced activation and conditions in which closures deactivated the device. With the remaining individual, the effects of momentary activation by switch closure were compared to timed activation. Data were collected with a combination of data collection technologies. Whole-session rate and response-duration data provided some evidence of learning in all participants. Data also were converted into within-session cumulative graphs, which aided data interpretation. The results demonstrate advances in assessment of learning in people with profound multiple impairments. Copyright © 2003 John Wiley & Sons, Ltd.