In order to examine the effectiveness of a token economy program, sheltered workshop attendance of chronic schizophrenics was investigated under the following conditions: response cost versus no response cost consequences for nonattendance; response cost and noncontingent token payment; and restricted versus unrestricted mobility during the no response cost condition. The greatest attendance occurred during the response cost conditions, while the least occurred during the no response cost, unrestricted mobility condition. Workshop attendance per se and the wages earned at the workshop were found to be reinforcing to many patients. (Author)