THE DATA FOR THIS STUDY CAME FROM A STRATIFIED PROBABILITY SAMPLE OF 560 TORONTO FAMILIES. THE SAMPLE WAS STRATIFIED ACCORDING TO WHETHER THE FAMILY RESIDED IN A DWELLING WHICH HAD MORE THAN ONE PERSON PER ROOM, (OF WHICH 273 DID). THE INDEPENDENT VARIABLES USED IN THE STUDY WERE OBJECTIVE AND SUBJECTIVE HOUSEHOLD CROWDING AND OBJECTIVE AND SUBJECTIVE NEIGHBOHOOD CROWDING. THE AGGRESSION MEASURES INCLUDED THREE DIMENSIONS: FAMILY AGGRESSION, PHYSICAL PUNISHMENT, AND OUTSIDE THE HOUSEHOLD AGGRESSION. MULTIPLE REGRESSION ANALYSIS WAS USED TO MEASURE THE IMPACT OF CROWDING ON THESE AGGRESSION MEASURES. THE EXTENT TO WHICH NEIGHBOHOOD INTEGRATION, QUALITY OF THE RELATIONSHIP BETWEEN THE MARRIED PAIR, ANXIETY LEVELS, AND ECONOMIC STATUS, INTERVENED IN THE RELATIONSHIP BETWEEN CROWDING AND AGGRESSION WAS ALSO INVESTIGATED. OBJECTIVE CROWDING CONDITIONS SHOWED LITTLE RELATION TO AGGRESSION WITH THE EXCEPTION OF THE ADMINISTRATION OF PUNISMENT TO CHILDREN. SUBJECT PERCEPTIONS OF HOUSEHOLD CROWDING, HOWEVER, WERE RELATED IN A SMALL BUT CONSISTENT WAY TO FAMILY AGGRESSION AND PHYSICAL PUNISHMENT FOR FEMALES, AND TO OUTSIDE THE HOME AND FAMILY AGGRESSION FOR MALES. GENERALLY, THE EFFECT OF CROWDING WAS NO MORE SEVERE WHEN PERSONS WERE UNDER STRESS THAN WHEN THEY WERE NOT. A THREE-PAGE LIST OF REFERENCES IS INCLUDED.