During the early to mid-1980s, the SADME (SA Department of Mines and Energy) Engineering Division collaborated with CSIRO Soils Division to jointly sponsor a major geomechanics project, investigating metropolitan Adelaide’s problem soils and near surface geology (Sheard and Bowman, 1996b). A small portion of that project involved an intense study of a stormwater-excavated erosion ditch and gully located within part of an undeveloped transport corridor at Modbury North. That area offered an opportunity to examine and test regolith profiles now exposed down to incipiently weathered rock, using geological, pedological, geomechanical and geochemical methods. Located on the present day Para Block, north of Grand Junction Road, the eroded area generally has a gentle SSE sloping landscape. Soils had formerly obscured subsurface materials and bearing horizon irregularities, that were only revealed when the erosion gully became incised to ~1 m to >4 m depth over a curving distance of >900 m. The gully exposed at least four unconformable palaeosurfaces, numerous soils and mosaics thereof, a variety of subsoil clays exhibiting varying seasonally active heave and/or collapse characteristics, and the underlying in situ weathered bedrock which consists of variably calcareous or dolomitic siltstone and slate. Most of the subsurface materials are also regolith carbonate - overprinted or are in part replaced by regolith carbonate. This site offered the opportunity to observe, due to excellent 3D exposures, a variety of geomechanical hazards that have potential civil engineering implications. Erosion gully topography-morphology and materials mapping, profile logging, sampling and testing revealed areas of black earth and rendzina soil profiles overlying gleyed smectite-rich Keswick Clay. This combination is very reactive to seasonal moisture variations, producing severe desiccation cracking in dry months and significant surface heave in wet months. The gully exposed weathered Neoproterozoic fine-grained Saddleworth Formation bedrock, and importantly, revealed three spaced erosional incisions that are Keswick Clay infilled, these subsequently being fully depth tested to >8 m by drilling. The incisions align orthogonally to the current surface slope, denoting a palaeolandscape with an earlier westward slope, and thus hinting at progressive episodic hinge movements of the Para Block during Late Quaternary millennia. It is evident that a simple geomorphology and stratigraphy have, with time and weathering, yielded a far more complex subsurface regolith architecture, and these components have given rise to a complex pedological surface distribution pattern. The palaeovalley infill is primarily composed of smectite-rich and vermiculite-bearing clays (XRD mineral examination); this mineralogy, along with trace element and oxide geochemistry, indicates that the infill is most likely derived from weathered local bedrock. This remains an intriguing concept, as this mineral-chemical association implies that a very different weathering regime applied to the landscape evolution prior to, during and then post- palaeovalley infilling. Geomechanical laboratory testing of soil materials revealed that, of the four dominant soil types recognised (black earth, rendzina, terra rossa, red-brown earth), only black earth and its intergrade mosaics with rendzina soil exhibited demonstrable moisture-related volume change behaviour. However, it was found that the Keswick Clay substrate can be up to twice as reactive as its overlying black earth. Their combined swell-shrink behaviours can generate local geomechanical issues of significant size over a variety of scales, involving short and long-term effects. The black earth and Keswick Clay also react strongly to soil moisture solute leaching caused by surface water ponding and infiltration, or occasioned via leakage infiltration from buried service pipe damage. At Modbury North, black earths (± rendzina soil) and Keswick Clay, when occurring together, usually exhibit gilgai microrelief. Gilgai can heave seasonally, or, if previously equilibrated, can quickly reactivate heaving due to human-induced changes in their local environment. These soil textural traits commonly impose some design limitations, or require additional foundation costs for most built structures to which they form a substrate. This study emphasises that soil mapping alone cannot delineate the total swell-shrink potential of a locality having mixtures of highly reactive and relatively non-reactive soils and/or related substrates. Geomechanically, the presence of buried irregularities like palaeovalleys pose serious localised footings design challenges to existing and planned infrastructures. This is especially so where cryptic bedrock incisions are infilled with highly to extremely reactive Keswick Clay or its informally named equivalents. If specific geomechanical test properties such as very high to extreme Atterberg Limits and high to extreme swell-shrink capacity are indicated for soils beneath proposed building sites, they are clear warning signs for later failures of standardised footings. The presence of low density, potentially collapsing regolith carbonate accumulations (>300 mm thick, >10 square metres in area, and with a bulk density of <1.5 t/cubic m) would also require additional design modifications and subsequent ongoing environmental conditions maintenance, to avoid significant footing failures due to sudden, irregular and irreversible surface collapse (by up to 20% of the original thickness). Modbury North’s overall geo-complexity and geomechanical hazard potential were not evident to the prime authors at first inspection, but those emergent aspects as described herein well demonstrate an important warning to suburban councils, state planners and developers, of how important it is to apply judicious care when proposing development and infrastructure on available lands over the elevated sloping Para Block.