The extent to which behavioral interventions achieve clinical significance frequently rests on whether gains made during an intervention transfer across settings and maintain over time. Critical to this measure of significance is the extent to which manipulation of variables in one setting affects responding in other settings. In the present study a method of obtaining stimulus control of upright walking posture used with one severely retarded adult was described and the effects that varying antecedent and consequent variables in a training setting have on upright posture in a transfer setting was examined. Results indicated that improved posture in the transfer setting was functionally related to manipulation of training-setting variables. Maintenance of posture gains in the transfer setting was excellent across a 3-month follow-up phase.