The thermal history of the Ardeche paleo-margin (southeastern France) has been reconstructed using various analytical data from minerals and organic matter and conventional interpretation. These methods comprise Tmax and vitrinite reflectance measurements on organic matter, determination of the smectite content in mixed-layer illite/smectite, fluid-inclusion microthermometry in quartz, barite, dolomite, and anhydrite, and fission-track analyses in apatite. The Balazuc (BA1) drill hole, 1730 m deep, intersects mainly Jurassic carbonate rocks, Triassic sandstones, and sulfate-rich claystones and ends in Carboniferous rocks. The present bottom-hole temperature is 70 degrees C. Other drill holes, less than 6 km away, were studied when necessary. One of the main objectives of this paper is to assess the agreement between the different geothermometric methods, and to check their limitations and convergence when used on carbonate- and sandstone-dominated reservoirs. In the Triassic sandstones, primary aqueous inclusions in anhydrite cement and at the boundary between the detrital quartz grain and the quartz overgrowth are one-phase brine inclusions entrapped below 70 degrees C. The homogenization temperature determined on the later barite cement is about 100 degrees C. Aqueous two-phase inclusions in healed microfractures in anhydrite are present at the base of the drill core near major faults. Temperatures as high as 210 degrees C recorded for these fluid inclusions could be associated with episodic fluid injections from below through the fault zones during the Early Jurassic. The fluid inclusions in dolomite, most of them reequilibrated during burial, indicate a trapping temperature of 130-145 degrees C at 1600 m and a temperature decrease towards the surface. This value is in good agreement with the MPTB (maximum paleotemperature of burial) method on organic matter, which converges to a maximum burial temperature of 130 degrees C at the same depth. The apatite fission-track data agree with these temperatures and indicate that the temperature decreased below 120 + or - 10 degrees C during the Eocene. These paleotemperatures, much higher than the present ones, can be explained by the erosion of 1900 m of mainly Cretaceous sediments. This interpretation implies a high average sedimentation rate during the Cretaceous, in agreement with data determined on a regional scale. A time-temperature evolution is proposed for the formations present in the Balazuc drill hole and the eroded sedimentary rocks. The application of the dual reaction model of the transformation of smectite to illite (Velde and Vasseur 1992) shows that the data are in good agreement with the kinetic model. However, the vitrinite reflectance variation trend with depth, predicted from the Burnham and Sweeney model (1989), is different from the data variation. No definitive explanation can be proposed for the failure of maturity assessment through vitrinite reflectance in the BA1 drill hole.