Well integrity is affected by two major issues: design and the underground environment. Anticlinales Perales Sur (APS) oilfield is located 30km to NE of Las Heras city (Santa Cruz) in Golfo San Jorge Basin (GSJB). In this oilfield an integrity well inventory reveals that 17% of wells are collapsed. No major reasons for collapsing were identified in well design, therefore the analysis was focused in the underground environment. APS oilfield is at the stage of waterflooding project and this project involves only Fm. Bajo Barreal. This continental unit has sandbodies product of fluvial deposition which are separated by 20 to 30m of muddy sections analyzed as fluvial plains.A 50% of collapses occurs in muddy sections very near to the contacts with sandbodies, 39% are direct in sand sections, and finally a 11% could not stablish. The spatial distribution of collapses do not correspond to any kind of range of depth, but also it is not related to main or secondary faults. That is why it is necessary to understand the geomechanics behavior of reservoirs once they are flooding. Some authors postulated that a heterogeneous reservoir when is injected expand unequal and exerts very different uplift in overlayers that it can affect mechanic stability of the area. This geomechanic scenario might be even more complex if we take in account an multilayer reservoir as those of GSJB. Superficial features as fissures oriented N-S and uplift near collapsed well locations might understand as a consequence of underground geomechanics instability. Several cases of land motion related to subsurface fluid injection has been documented by studies using Interferometric Synthetic Aperture Radar (InSAR). This methodology was applied in APS and it let detect variations in the oilfield surface. In summary, the monitoring of water injection and InSAR data in the area will let evaluate if the hypothesis that the fluid injection in a heterogeneous multilayer reservoir might create a geomechanic stress which it produces well collapse, land montion and surface features detected.