It is important to grasp the significance of and stress relaxation behaviors of the paving materials when considering the non-traffic cracking of asphaltic pavements1), 2). The mechanical properties that concentrate on the stress relaxation of various paving mixtures have already been reported in the reference 1) and 2).Currently, there are few reports in which the coefficient of expansion, α, of asphalt cements and epoxy resinsh as been studied. However, there are only few papers that deal with of practicalp aving mixtures over a wide range of temperature.In this paper, we shall discuss of the paving materialso n the temperaturest he pavement will encounter during serviece. The term thermal characteristics in this paper means not only α but also the glass transition point relating to a specific temperature and to others.The paving materials employed in this experiment are a straight asphalt cement, modified asphalt cement, epoxy asphalt cement and epoxy resin; and each mixture is obtained by mixing a binder with the aggregates.The degree of expansion and contraction of a specimen was measured using a contact strain gauge. The cooling and heating rates of the environmental room used for laboratory experiments are shown in Table 1. Five minutes after ascertaining the temperature of a dummy specimen pre-set with a thermocouple, the measurements were carried out for thirty minutes.As is generally known, the specimen expanded with rising temperature and contacted with falling temperature (Figs. 1-8). The α's of the straight asphalt cement, modified asphalt cement, epoxy asphalt cement and epoxy resin were about 1.40×10-4(cm/cm/°C), 1.24×10-4(cm/cm/°C), 1.17×10-4(cm/cm/°C), 1.04× 10-4(cm/cm/°C), respectively. The α of each mixtures was smaller by about one order than its own binder because of the very small α of the aggregates, for example, the order of 10-6, and the total weight of aggregates was 94.2% in the mixture5), 6). These values are summarized in Table 2.The glass transition point Tg, was revealed in the expansion and contraction vs. temperature curve of the paving mixtures. The α of the mixtures was larger at temperatures above Tg and smaller at temperatures below Tg. For example, α of epoxy asphalt mixtures was 2.35×10-5(cm/cm/°C) at temperatures above Tg and 1.50×10-5(cm/cm/°C) at temperatures below Tg. Table 3 summarizes the Tg's of the four kinds of mixtures.In general, the correlation between the glass transition point Tg and the specific temperature Ts can be expressed as fbllows11). Ts=Tg+50(±5). Ts is calculated from the WLF equation8), 9) using the relation between the shift factor and temperature. Slight difference were observed in the relation between Ts and Tg obtained from our study. We consider these difference are probably caused by experimental errors and by the constants used in the WLF equation. However, the order of the difference in Tg and Ts was not of practical significance.We also measured expansion and contraction of an experimental pavement of epoxy asphalt mixtures (Fig. 9).