Primary salt welds form at the base of minibasins in response to complete evacuation of autochthonous salt. Analytical and numerical models suggest it is difficult to completely remove salt from a weld by viscous flow alone, which is especially true in multilayered evaporites, within which flow is likely heterogeneous due to lithologically controlled viscosity variations. Welds are important in the hydrocarbon industry because they may provide a hydrodynamic seal and trap hydrocarbons, or may allow transmission of fluids from source to reservoir rocks. Few papers document the subsurface expression of welds, principally because they have not been penetrated by wells or because the associated data are proprietary. We use 3D seismic and borehole data from the Santos Basin, offshore Brazil to characterize the geological and geophysical expression of a primary weld associated with flow of Aptian salt. The seismic data that we evaluated suggested that, locally, presalt and postsalt rocks are in contact at the base of an Upper Cretaceous minibasin, implying that several apparent welds, separated by low-relief salt pillows, are present. However, borehole data indicated that 22 m of anhydrite, carbonate, and sandstone are present in one of the welds, indicating that this and other welds may be incomplete. We find that seismic data may be unable to discriminate between a complete and incomplete weld, and we suggested that, during the subsurface analysis of welds, the term apparent weld is used until borehole data unequivocally proves the absence of salt. Furthermore, we speculate that preferential expulsion of halite and potash salt from the autochthonous layer during viscous flow and welding resulted in the formation of an incomplete weld, which, when compared with the initial autochthonous layer, is volumetrically enriched in nonevaporite lithologies and relatively viscous evaporite lithologies (anhydrite). The composition and stratigraphy of the autochthonous layer may thus dictate weld thickness and seal potential.