The individualized nature of instructional programs for the mentally handicapped often makes group designs inappropriate in adapted physical activity research. Single-subject time-series designs are suitable for use in investigating the acquisition, maintenance, and generalization of motor skills when the research involves small numbers of subjects. These designs require the collection of data before, and during or after treatment. Three single-subject time-series designs are described and illustrated with data from studies in the PREP Play Program, an instructional program for young mentally handicapped children at the University of Alberta. The simple time-series design has severe limitations for use as a research tool, but is appropriate for use by teachers or practitioners who are monitoring previously tested treatments in physical activity programs. The repeated time-series or reversal design can be used to investigate the maintenance or generalization of effects after treatments are withdrawn. The multiplebaseline design is recommended for researchers or practitioners who wish to assess the effects of instructional programs on different subjects or different dependent variables.