Recent studies have shown large increases in vehicle emissions when the air conditioner (AC) compressor is engaged. Factors that affect the compressor-on percentage can have a significant impact on vehicle emissions and can also lead to prediction errors in current emissions models if not accounted for properly. During 1996 and 1997, the University of California, Riverside, College of Engineering-Center for Environmental Research and Technology (CE-CERT) conducted a vehicle activity study for the California Air Resources Board (CARB) in the Sacramento, CA, region. The vehicles were randomly selected from all registered vehicles in the region. As part of this study, ten vehicles were instrumented to collect AC compressor on/off data on a second-by-second basis in the summer of 1997. Temperature and humidity data were obtained and averaged on an hourly basis. The ten drivers were asked to complete a short survey about AC operational preferences. This paper examines the effects of temperature, humidity, refrigerant type, and driver preferences on air conditioning compressor activity. Overall, AC was in use in 69.1% of the trips monitored. The compressor was on an average of 64% of the time during the trips. The personal preference settings had a significant effect on the AC compressor-on percentage but did not interact with temperature. The refrigerant types, however, exhibited a differential response across temperature, which may necessitate separate modeling of the R12 refrigerant-equipped vehicles from the R134A-equipped vehicles. It should be noted that some older vehicles do get retrofitted with new compressors that use R134A; however, none of the vehicles in this study had been retrofitted.