IN CONTRAST TO STUDIES OF POLICE DISCRETION THAT FOCUS ON INDIVIDUAL BEHAVIOR, THIS STUDY ATTEMPTS TO LOCATE THE INTERACTION CONTEXT THROUGH THE DIMENSIONS OF SOCIAL OCCASIONS. GOFFMAN DEFINES A SOCIAL OCCASION AS AN AFFAIR, UNDERTAKING, OR EVENT BOUNDED BY TIME AND PLACE WITHIN WHICH MANY SITUATIONS AND GATHERINGS FORM, DISSOLVE, AND REFORM. OCCASIONS ARE NOT AS PERMANENT OR AS STRUCTURED AS THE LARGER SOCIAL ORGANIZATIONS WITHIN WHICH THEY ARISE. IN ORDER TO TEST THE RELATIONSHIP BETWEEN OCCASIONS AND POLICE MOBILIZATION, DATA WERE GATHERED OVER A 5-MONTH PERIOD. A TOTAL OF 291 POLICE MOBILIZATIONS WERE OBSERVED IN A SMALL CITY'S POLICE DEPARTMENT. OBSERVATION SCHEDULES WERE DEVELOPED FOR RECORDING DATA; OCCASIONS ON THE SCHEDULE WERE GIVEN THESE DIMENSIONS: LOOSE, OPEN, SCHEDULED, UNSCHEDULED, CLOSED, TIGHT, AND FOCUSED AND UNFOCUSED. FORMS OF INTERACTION WERE HYPOTHESIZED TO BE DEPENDENT ON THE SOCIAL OCCASION IN WHICH THE POLICE AND CITIZENS FOUND THEMSELVES. FINDINGS VERIFIED THAT LOOSE AND OPEN OCCASIONS WERE MORE LIKELY TO LEAD TO POLICE INTERVENTION SINCE THEY WERE OCCASIONS WHEN FEELINGS COULD GET OUT OF HAND AND ANTAGONISM COULD ERUPT. THERE WAS ALSO A SIGNIFICANT RELATIONSHIP BETWEEN THE TYPE OF OCCASION AND THE LIKELIHOOD OF POLICE INTERVENTION, THUS LEADING TO SPECULATION THAT CRIME AND DISORDER ARE RELATED TO FORMS OF SOCIAL OCCASIONS IN SPECIFIC DIMENSION. (AUTHOR ABSTRACT MODIFIED--MJW)