Federal income support to persons with alcohol and drug related disabilities was ended in 1997. The argument for ending the programs was that recipients were using their benefits to purchase drugs and alcohol. This supposedly led to the "check effect," an increase in psychiatric emergencies in American communities in the days following the receipt of disability benefits. We test two hypotheses implied by this argument. The first is that psychiatric emergencies are elevated in the fourth through eighth day of the month. The second is that the excess of emergencies in these days was significantly reduced when benefits were ended. The tests are based on 35,500 psychiatric emergencies in San Francisco, California occurring over 1,551 days. Results support the first hypothesis but not the second. The implications are that there is a general check effect and that it was not reduced by ending benefits to persons with drug and alcohol related disabilities.