Saltmarshes in subtropical Queensland, Australia, are assumed to provide habitat for fish, yet there has been no evidence as to whether fish actually occur on the intertidal flats that constitute the main area of marshes. Nor has there been any assessment of differences in fish use of vegetated habitat and the extensive unvegetated saltpans that occur naturally and because of human activities. Fish were sampled from the intertidal flats of 2 saltmarshes in subtropical Queensland using floorless, buoyant pop nets (25 m 2 ) on sets of spring high tides in summer (during day) and winter (at night). A total of 23 species was collected, with densities ranging up to 0.46 fish m -2 . Catches were dominated numerically by small, estuarine-resident species such as glassfish (Ambassidae) and gobies (Gobiidae). Economically important species such as bream (Sparidae), whiting (Sillaginidae) and mullet (Mugilidae) also contributed substantially (up to 92%) to the catch. Fish assemblages were compared in adjacent patches of intertidal vegetated and unvegetated habitat, at distances varying from 3 to 413 m onto the marsh. Fish densities were highly variable among patches on a marsh, and also varied between sampling periods and between marshes. Species richness and the densities of common species differed little between vegetated and unvegetated habitats. Two species, Mugilogobius stigmaticus and Atherinomorus ogilbyi, had higher densities in vegetated habitat, but only at 1 sampling time and marsh. Fish occurred at all distances onto the marsh. Species richness and the densities of some species were positively related to water depth and negatively related to distance onto the marsh at one, but not the other marsh. This study provides the first evidence that, despite the relatively low duration and frequency of inundation of marshes, there is widespread use by fish of intertidal saltmarsh flats in Australia.