Information from experimental settings of gill nets reported in the literature and data taken from commercial fishermen in Florida from 1973 to 1981 were used to analyze gill net selection on Spanish mackerel, Scomberomorus maculatus, king mackerel, S. cavalla, and bluefish, Pomatomus saltatrix from Florida and Serra Spanish-mackerel, S. braziliensis, from Brazil. The information on Spanish mackerel and bluefish obtained from the commercial gillnet fishedes did not further our understanding of selectivity over that produced by experimental nettings except in defining girth-length relations. The commercial fisheries data appeared to reflect mostly the sizes of fish that were abundant at the time of capture ratner than the effects of selectivity. We did use commercial fisheries data for describing selectivity of king mackerel, however, because we had other estimates of the size compositions of the populations to adjust the gillnet distributions for unequal numbers of fish in the length intervals. Selectivity was evaluated under the assumptions that: (1) the selectivity curve would take the form of a normal frequency distribution; (2) the efficiencies of two nets with different mesh sizes would be similar for fish of their respective lengths; and (3) the standard deviations of the distributions for two different mesh sizes would be equal. Under these assumptions the computed mean selection lengths in relation to mesh size and species- were: Spanish mackerel - 6.3 cm stretched mesh, 30.8 cm fork length; 7.0 cm SM, 33.9 cm FL; 7.6 cm SM, 37.0 cm FL; 8.2 cm SM, 40.1 cm FL; 8.9 cm SM. 43.2 cm FL; and 9.5 cm SM, 46. 3 cm FL. Bluefish - 6.3 cm SM, 28.5 cm FL; 7.0 cm SM, 31.4 cm FL; 7.6 cm SM, 34.2 cm FL; 8.2 cm SM. 37.1 cm FL; and 8.9 cm SM, 40.0 cm FL. King mackerel - 12.1 cm 5M, 92.-1 cm FL. Serra Spanish mackerel - 6.0 cm SM. 42. 1 cm FL; 8.0 cm SM. 43.6 cm FL; and YO cm SM, 55.2 cm FL.