Zhou et al. (2012) proposed that longitudinal dunes in the Qaidam Basin, China, formed like yardangs: by erosion into sediment that was not deposited by those dunes. Because erosion occurs on the upwind fl anks of most migrating dunes (Rubin and Hunter, 1982, 1985), the key to demonstrating a yardang-like origin is to show that the dunes did not deposit the strata that they contain. Zhou et al. made this argument by proposing that: (1) The dunes have not deposited cross-strata in the past 810 yr. (2) Cross-bedding within the dunes was not deposited by the dunes on the present-day land surface, but rather by older dunes that had a different morphology. (3) The present dunes are a later generation, “most likely of erosional origin similar to yardangs with orientations controlled by strikes of joints,” (p. 1147). (4) Rates of defl ation in the dune fi eld have been extremely high for the past 810–2440 yr. This commentary reviews these conclusions, reviews contradictory observations, and considers alternative interpretations.