Climate variations in both space and time can impact mountain topography and drainage-basin morphology through a variety of surface processes. This is illustrated dramatically in the Ordos Block of northern China, which encompasses the main body of the Loess Plateau in the southeast and the Mu Us Desert in the northwest. The plateau is known for severe soil erosion by fluvial processes; whereas the desert is dominated by aeolian processes. Here we characterize the spatial transition in landscape relief and drainage network integration of the middle Ordos Block, corresponding to a precipitation gradient across the region, that is otherwise homogeneous in terms of its tectonics and bedrock. Differences in landscape relief and drainage network integration occur over a short range in precipitation rate, and outside this transition zone, landscape morphologic metrics appear to be relatively insensitive to precipitation. Where precipitation rates are greater than -420 mm/yr, landscapes are characterized by integrated drainage networks with relief over 100 m within a 1 km radius. With precipitation rates lower than -380 mm/yr, the landscapes tend to have lower relief, discontinuous channels and hundreds of playa lakes indicating inefficient runoff. Combined with stratigraphic records, we posit that continuous drainage networks in the semi-arid to arid transition zone expand in wetter periods, and shrink and fragment when precipitation falls below a threshold. Today the Ordos Block is experiencing the transition from a drier (glacial) stage to a wetter (interglacial) stage, with higher precipitation rates over the past -12 thousand years. It follows that fluvial processes likely have increased in efficiency and are currently advancing into regions to the northwest formerly dominated by aeolian processes. Our findings suggest that a component of the severe sediment erosion of the Loess Plateau is a transient feature locally associated with glacial-interglacial climate change and the transition from aeolian- to fluvial-dominated surface processes.& COPY; 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons .org /licenses /by /4 .0/).