To assess cool-roof benefits, the temperatures, heat flows, and energy uses in two similar single-family, single-story homes built side by side in Fresno, California were measured for a year. The "cool" house had a reflective cool concrete tile roof (initial albedo 0.51) with above-sheathing ventilation, and nearly twice the thermal capacitance of the standard dark asphalt shingle roof (initial albedo 0.07) on the "standard" house. Cool-roof energy savings in the cooling and heating seasons were computed two ways. Method A divides by HVAC efficiency the difference (standard cool) in ceiling + duct heat gain. Method B measures the difference in HVAC energy use, corrected for differences in plug and window heat gains. Based on the more conservative Method B, annual cooling (compressor + fan), heating fuel, and heating fan site energy savings per unit ceiling area were 2.82 kWh/m(2) (26%), 1.13 kWh/m(2) (4%), and 0.0294 kWh/m(2) (3%), respectively. Annual space conditioning (heating + cooling) source energy savings were 10.7 kWh/m(2) (15%); annual energy cost savings were $0.886/m(2) (20%). Annual conditioning CO2, NOx, and SO2 emission reductions were 1.63 kg/m(2) (15%), 0.621 g/m(2) (10%), and 0.0462 g/m(2) (22%). Peak-hour cooling power demand reduction was 0.88 W/m(2) (37%). (C) 2014 Elsevier B.V. All rights reserved.