The Maracaibo Basin is a 50,000-km(2) (19,305-mi(2)) topographic depression bounded to the east and south by the Merida Andes of Venezuela and to the west by the Sierra de Perija of Venezuela and Colombia. Both uplifted mountain blocks expose Paleozoic basement rocks and Mesozoic-Cenozoic carbonate and clastic rocks that were mainly folded and thrusted by regional shortening in the Paleogene and late Neogene. Using geologic maps, seismic reflection data, and wells from the steep mountain front areas, we test different structural models of how late Paleogene-Neogene convergence is accommodated along subsurface faults and folds at the mountain fronts. Seismic imaging of the deep (> 5-km; > 3.1 mi) structure of both Maracaibo mountain fronts shows basinward-dipping monoclines with stratal dips ranging from 20 degrees to overturned and an almost complete lack of faulting in the basin-edge monocline or in adjacent, horizontally bedded strata of the Maracaibo Basin. Seismic data reveal the presence of one or more triangle zones at depth along both the Merida Andes and Sierra de Perija that exhibit characteristic thrust-wedge geometries. The lower part of the wedge is defined by imbricate thrust faults dipping beneath the mountain block and involving the Paleozoic basement. The upper part of the wedge is defined by a basinward-dipping thrust associated with fault-propagation folds at the surface and an overlain basin-edge monocline. The creation of the steep to overturned dips of the monocline is attributed to the effects of 6-10 km (3.7-6.2 mi) of shortening along the lower zone of imbricated thrust faults. This 6-10 km (3.7-6.2 mi) of shortening, calculated from triangle zones of both the Sierra de Perija and Merida Andes, is significantly less than regional estimates from 12 to 60 km (7.4 to 37 mi) of shortening inferred by previous workers from regional balanced cross sections that assume low-angle thrust-type geometries. We propose that a pop-up style of deformation related to the inversion of Jurassic rift features may be a more realistic interpretation of the convergent fault systems that have controlled uplift of the Merida Andes and Sierra de Perija. Inversion of relatively steep, basement normal faults at the edges of and within both ranges may explain lesser amounts of observed shortening.