Unaltered salt marsh pools in southern New Jersey were sampled during the summer over a number of years, with a variety of techniques, to compare fish species and size composition relative to sampling gear type and to enhance our understanding of marsh pool fishes. These pools were dominated by a few species (Fundulus heteroclitus, Cyprinodon variegatus, Menidia beryllina, Lucania parva, and M. menidia made up 98.3% of all fish [n = 33,731] collected). However, species composition clearly varied with sampling technique, with some species common in multiple gears, e.g., F. heteroclitus in quatrefoil traps (38.6% of total number) and wire mesh traps (35.5%), and C. variegatus collected in quatrefoil traps (47.6%) and mini-seine (30.9%). Other species were most abundant in selected gears, e.g., M. beryllina (83.8%), L. parva (72.6%), and M. menidia (99.6%) in quatrefoil traps. Size composition, which included young-of-the-year and adults for most species, varied with sampling technique and species as well. In all of the above, we cannot rule out the possibility that annual variation influenced species composition and abundance; however, given the stability of these measures in other informal observations, we are convinced that most of the variation is due to sampling technique. Continued studies are relevant because marsh pools have been eliminated by a variety of practices, but are also being created as the result of some mosquito-control techniques and for restoration purposes.