The 'hyporheic refuge hypothesis' predicts that the hyporheic zone, the saturated sediments below and alongside rivers and streams, is a refuge from the scouring effects of spates for many aquatic invertebrates including water mites. We tested this hypothesis in two lateral gravel bars and two riffles in a subtropical Australian river by collecting water mites from the hyporheic zone at two depths (10 and 50 cm) at two 'pre-flood' sampling times before experimentally diverting water through the sites for 14 h to simulate a spate. Taxon richness of mites washigh (46 taxa) and dominated by the Prostigmata, with nearly half the species being new to science. Oribatids were also common at the four sites. Samples were collected twice during each 'spate', and again soon after flow was returned to normal. The experimental spate induced changes in the strength and even direction of subsurface-surface water exchange; however, these changes seldom persisted after the experiment, nor after a subsequent natural spate. The hyporheic refuge hypothesis was not supported by our water mite data. Neither during nor shortly after the experimental spates did we find more epigean (surface-dwelling) water mites in downwelling zones where surface streamwater enters the hyporheic zone, demonstrating that these mites were not using the hyporheic zone as a refuge at these locations. There was also no evidence for a 'wash out' effect, because hyporheic mitedensities did not significantly decline late in the spate. Our data indicate that floods of the low magnitude simulated in this study apparently do not pose a lasting disturbance for hypogean water mites. The fact that the same response was found at four sites indicates that the hyporheic refuge hypothesis may not always be an appropriate explanation for rapid post-flood recolonisation. Possibly, the use of the hyporheic zone as a refuge from floods may be dictated by the strength of the disturbance and substrate composition and stability.