The formation of iron duricrusts has been the subject of much debate. First hypotheses emerged in the 1950's, when they were first described in Africa, Australia or India. Ferruginous duricrusts form as hard iron layers in tropical and subtropical environments with strong climatic seasonality. The accumulation of iron oxyhydroxides during the wet season and precipitation during the dry season is crucial for duricrust formation. Their extremely high iron content makes them more resistant to mechanical weathering. This resistance to erosion explains why they often cap and protect landscapes. Some have argued that duricrusts form at the surface but others have described them buried at depth in the regolith, notably in Africa, India or South America. This has led to 2 distinct hypotheses: the first one relies on the lateral hydraulic transport of iron and on the vertical beating of the water table most likely in a stable environment, whereas the second one relies on the vertical, in situ leaching of elements and resulting compaction and surface lowering. According to the 1st hypothesis, iron hydroxides are transported from surrounding regions through the water table to accumulate in one specific area where Fe3+ will precipitate during the dry season. Through time, this leads to the formation of a duricrust between the water levels of the wet and dry seasons. In this case, required elements are brought from a large source area. This model is compatible with the formation of duricrusts at depth. The 2nd hypothesis implies a genetic link between the underlying parent rock, the lateritic profile and the duricrust. During weathering, soluble elements are progressively leached from the parent rock leaving only the iron oxides, which concentrate at the surface to form nodules and ultimately, a duricrust. This model explains why duricrusts are found at the surface and the genetic link with the bedrock. There is however no consensus on the conditions under which either of these hypotheses prevail. We, therefore, created 2 separate models for the formation of duricrusts that represent the two hypotheses. By running a large number of simulations, we used the models to demonstrate that the hypotheses lead to different scenarios of duricrust formation, especially we show that the stability of a landform is a discerning element between the 2 models.