There is growing awareness that patterns ofhabitat use by animals cannot be isolated from issues ofscale. Recently, techniques have been devised which allowempirical testing of hypotheses related to the effectsof spatial scale on habitat use. We used spatially explicitstatistical procedures to examine the roles of scale andhabitat arrangement in determining fish distribution patternsin a reach of an Australian lowland river. Nativegolden perch and introduced common carp were trackedday and night using radiotelemetry over a 10 day periodand their distributions compared with mapped habitatvariables. Golden perch were significantly associatedwith depth, current velocity, substratum and cover atlarger scales of analysis both day and night, and at smallerscales during the day. At night, however, associationsbetween golden perch and habitat variables were generallymuch weaker at small scales. Common carp weregenerally not significantly associated with habitat variablesat larger scales of analysis, whereas at smallerscales they were associated with depth, current velocity,substrate and cover. Associations were generally strongerduring the day than at night. Our study emphasises theneed to consider scale-dependence in studies of fish-habitatassociations. Management and restoration of fishhabitat in lowland rivers should be based on an understandingof habitat use at scales relevant for the fish ratherthan at scales arbitrarily defined by humans.