Abstract The Soria Basin is a rhomb graben with borders that trend N60°E and N50°W. It was formed during the Late Jurassic -Early Cretaceous (Wealdian), when as much as 8 km of fluvially dominated deltaic strata accumulated in it. This sedimentary fill has been divided into five cyclothems, of which the lower four are discussed in this paper. Within the basin, a N50°W-trending, 50-km wide, syn-sedimentary syncline developed in the basin fill. This syncline was related to extensional tectonics, and to the formation of a half graben in Paleozoic basement overlain by competent Jurassic and incompetent Triassic strata. Within the basin fill, the extensional deformation produced microstructures (stylolites, calcite tension gashes, and quartz dikes) with a coherent basin-wide partem. The depocenter migrated with time from the southeastern comer of the basin during deposition of cyclothems I and II, to the northeastern corner during formation of cyclothems III and IV. High heat flow, related to crustal thinning in the area of greatest subsidence, led to metamorphism within the sediments. Conditions of metamorphism were a maximum temperature of 420° C, a temperature gradient of 100-150° C/km and pressures between 1-3 kb. At the same time, compressional deformation (N30°W-trending folds and associated cleavage) was induced along the southeastern margin of the basin, and erosion (uplift) occurred outside the basin. Our interpretation of the geometry, sedimentation, tectonics, and thermal evolution of the Soria Basin is based on mathematical models and microtectonic analogue models of a releasing solitary overstep. In such models, stress/strain deviations and accumulations predict vertical motions (subsidence and uplift), and the geometry of structures in areas of extension (secondary normal faults, tension gashes) and in areas of compression (folds, cleavage). Most of the field data collected inside and outside the basin are consistent with a model of a releasing overstep along N60°E-striking sinistral strike-slip faults. The proposed releasing overstep model differs from classical models of strike-slip basins by (1) taking into account stress/strain related to basin development, (2) explaining migration of the depocenter with time, and (3) predicting the geometry of secondary faults.