The incidence of alcohol in signalised intersection accidents is examined and signal design and operation measures with some practical potential for improving safety are identified. Evidence from a review of the driver performance literature indicates that alcohol impairs a number of driver performance functions upon which an increased demand is imposed by intersection signals. Accident data analysis was carried out on 7,402 reported casualty accidents which occurred at signalised intersections in the Melbourne metropolitan area, between 1977 and 1980, inclusive. In about 39 per cent of these accidents, blood alcohol readings for at least one of the drivers involved were available from hospital testing procedures. The principal findings of the analyses of intersection signal accidents were: 1) right- turn- against, cross- traffic and rear- end accidents were the predominant accident types; 2) 86 per cent of alcohol- related accidents occurred between 6 pm and 6 am; 3) alcohol was over- represented in rear- end accidents; 4) alcohol was over- represented in the general category of single- vehicle/ loss- of- control accidents; and 5) alcohol was under- represented in cross- traffic and right- turn- against accidents. Only a limited number of signal design and operation countermeasures were identified for these accident types. The principal countermeasures were improved lane delineation and more conspicuous signal displays. The thesis was submitted in fulfilment of one- eighth of the requirements for the degree of Master of Engineering Science (Transport) in the Department of Civil Engineering, Monash University, September, 1982. (TRRL)