Direct counts of fish obtained by night spotlighting were compared with species composition and population estimates obtained from three-pass electrofishing obtained across 29 sites along small clear streams in the Otago region of New Zealand. The influence of habitat variables on the relative efficiency of each method was also examined. The same seven species of fishes were identified by both methods. Juvenile brown trout Salmo trutta were the only species present in sufficient numbers to allow comparison of abundance estimates using the two methods. A total of 777 brown trout were counted by spotlight and 803 brown trout were caught using electrofishing. Estimates of abundance obtained by spotlighting reflected population estimates obtained by three-pass electrofishing across most habitats. Electrofishing produced higher population estimates relative to spotlighting in fast-flowing turbulent riffle habitats, whereas counts obtained by spotlighting tended to be higher relative to electrofishing in slow-flowing pool habitats. The results suggest that spotlighting is an effective method for assessing fish composition and brown trout abundance in small clear water streams, although the extremes of water velocity may influence efficiency of both spotlighting and electrofishing.