A prototype, boat-mounted electrofisher capable of operation in estuarine waters (where electrical conductivities often exceed 20 000 µS cm−1) was assessed. Electrofishing was compared to fyke and mesh netting in four riverine estuaries and to seining in a lagoonal estuary (consisting of a series of brackish coastal lakes separated from the sea by a barrier system of sand dunes). Fish assemblage composition, length distributions and the probability of detecting ecological fish guilds (relating to estuary use, position in the water column and body size) were compared among gears. The assemblage composition of electrofishing samples differed from those of fyke nets in all riverine estuaries and from mesh netting in two. The assemblage composition of electrofishing samples differed from those of seining in structured seagrass habitats of the lagoonal estuary. When all species were pooled, the electrofisher sampled a broader range of lengths than either fyke or mesh netting in riverine estuaries or seining in lagoonal estuaries. The bias of electrofishing and netting towards particular species and size classes affected the probability of detecting some ecological guilds, highlighting the potential implications of gear choice on understanding estuarine ecological function. The detection of guilds varied with gear type and environmental conditions, including stratification, water depth and surface electrical conductivity. Assessments with the aim to characterize the structure of fish assemblages will benefit from the use of multiple gears. Electrofishing shows immense promise for discretely sampling highly structured habitats to test hypotheses about their use.