An examination was conducted by a teacher who was interested in how the responses of three multihandicapped students (7-8 years old) differed in one-to-one and one-to-three instructional arrangements given (1) the physical limitations of the students; (2) an activity that demanded much physical prompting; and (3) limited instructional resources. Data were collected on the motoric responses made by each student during a 45 minute snack session. In the one-to-three arrangements, the teacher instructed the students concurrently. In the one-to-one arrangements, the teacher instructed each student consecutively. Thus, during the one-to-one arrangements, a student received one-to-one instruction for a portion of the 45 minute session and was wheeled to a free time area for the remaining portion. Results indicated that the students made more task-relevant responses and fewer counterproductive responses during the one-to-three arrangement than during the one-to-one arrangement when the data from the free time portions were included in the analysis. Discussion focused on the merits and limitations of one-to-one arrangements and the implications of these findings for the design of future research efforts. (Author/CL)