SynopsisThe strip count method of determining fish densities was investigated for a small lake in a series of inshore sites that differed in their physical and biological characteristics and in the communities of fishes present. Overall total numbers of fish, number of species, ratio of commonest to total species, and 2nd commonest to total species, were investigated. Counts were made during time of the feeding peaks.Repeatability of the method was tested by series of five traverses at lo-minute intervals and was good both for total numbers of fish and ratio of commonest species to total fish. However, it varied with the habitat type, being excellent for small semi-isolated weedy sites, good for gravel-bottom sites, but only fair in open sandy bays where the strip was laid in the middle of an extensive area of homogeneous habitat.Series of comparative counts made (a) between morning and afternoon on the same day and, (b) between successive mornings, to determine how constant populations were between these time intervals, proved to be within the same general range but were significantly different by the Mann Whitney U test at the D = 0.05 level.In small lakes the strip count method proves to have considerable potential in the study of habitat specializations, taxocene structures, and relative abundances.