The structures of macrobenthic communities that colonized sand-filled boxes in laboratory and field were compared to learn if laboratory-developed assemblages of animals used in toxicity studies realistically simulate those in nature. Laboratory boxes were colonized for eight weeks in spring and fall by planktonic larvae from continuously-supplied unfiltered seawater; field boxes, by naturally-occurring estuarine animals. Laboratory communities had larger numbers, but fewer species than field communities. One-fourth of the species found in field boxes also occurred in laboratory boxes. Species density, diversity, and dominance were similar in both environments in fall, but differed in spring. The degree of similarity of laboratory assemblage of benthic animals to the highly variable assemblages in the field indicates applicability of laboratory toxicity studies to nature.