Experimental flat roof systems were studied at an outdoor test facility. Heat flow rates and temperatures were measured with heat flow meters and thermocouples, and the values were recorded on magnetic tape at fixed intervals. The results were averaged over 24-h periods. Thermal resistances of systems and individual components were found using average net heat flow rates and temperature differences. Two analytical methods were used. In the first, the ratio of heat flow to temperature difference was calculated. This method was unreliable at low heat rates due to a bias in the results. In the second method, the slope of heat flow versus temperature difference was used as the basis for calculation. Thermal resistances of several pieces of insulation found by this method were compared with results obtained using laboratory equipment. The differences ranged from about 0.5 to 5 percent.