Abstract We present the results of field studies concerning the performance of sunscreens for reducing air-conditioning loads in single-family residences. We used data on houses (725) in the San Diego transition climatic region which had sunscreens installed during 1978 or 1979. The sample size was reduced by the following factors: failure by homeowners to respond to the questionnaire, houses which were not air-conditioned, failure to satisfy characteristics important for matching of houses, and criteria related to changes in occupancy. As the result, our control and test groups were reduced to 34 and 32 for the 1978 study and to 36 and 31 households for the 1979 study, respectively. For these small remaining sample sizes, the predicted small effects of sunscreens on total electricity consumption were obscured by variances in the field results. The observed large variances in the field results were produced, in large measure, by deliberate changes in thermostat settings as a response by homeowners to rapidly escalating electricity costs.