A technique for insular defaunation utilizing methyl bromide fumigation was evolved. Seven islands in Florida Bay, of varying distance and direction from potential zoogeographic source areas were censused exhaustively. The small size (45 ft. diameter) and ecological simplicity of these islands, which consist solely of red mangrove (Rhizophora mangle) trees with no supratidal ground, allowed the determination of all resident species. The completeness of the censuses was corroborated by chloropicrin fumigation. The breeding land fauna of these islands is composed almost exclusively of terrestrial arthropods, with 20-50 species usually present. Surveys of these taxa in the Florida Keys, with emphasis on mangrove inhabitants, were made during 1967. The seven experimental islands were defaunated in early 1967, and were then monitored for colonists for 17-20 man-hours every 18 days. Precautions were taken to prevent contamination by investigators and to insure recording of crepuscular and nocturnal forms. The colonization of six small mangrove islands in Florida Bay by terrestrial arthropods was monitored continu-ally for one year after defaunation. Both the observed data and climatic considerations imply that seasonality had little effect upon the colonization curves of species present vs. time. By 250 days after defaunation the faunas of all but the most distant island had returned to a state not unlike that on untreated islands in species number and composition, although population densities were still abnormally low. Early colonists included both weak and strong fliers, but the former, particularly psocopterans, were usually the first to produce large popula-tions. Among these same early invaders were the taxa displaying the highest extinction rates and also the greatest variability in species composition on the different islands. Ants, the ecological dominants of mangrove islands, were among the latest to colonize, but did so with the highest degree of predictability. The colonization curves plus static observations on untreated islands imply strongly that a dynamic equilibrium number of species exists for any island. The colonization curves are believed best explained as the product first of colonization involving little if any interaction and then a gradual decline as interaction becomes important, termi-nating in a lasting dynamic equilibrium. Equations are given for the early immigration, extinction, and colonization curves. Dispersal to these islands is predominantly through aerial transport, both active and passive. Extinction of early colonists is probably rarely related to biotic inter-action but rather to such physical factors as drowning or lack of suitable breeding site. As population sizes increase it is expected that competition and predation will become more important. Observed turnover rates showed wide variance , with most values between .05 and .50 spp./day. True turnover rates are probably much higher, with .67 spp./day the extreme lower limit on any island. A distinction is made between immigration rate (in spp./time) for an island and invasion rate (in propagules/time) for a species and an island. An analogous distinction is drawn between an island extinction rate and a species extinction rate (or intrinsic probability of extinction in a given time interval). It is claimed that the most objective definition for "propagule" is any animal or group capable of population increase under any conceivable circumstances. Immigration and island extinction curves are unique only if plotted against time, not against number of species. Most members of the Florida Keys species pool are obligate transients on small mangrove islands (often because of inadequate food or microhabitat) and those which can survive were present on the defaunated islands in such low densities that extinction was probably initially rarely due to interaction. Drowning of entire small populations prob-ably occurred frequently. A model for non-interactive colonization is discussed and its equilibrium Ŝ derived. It is shown that data from the defaunated Florida Keys can be interpreted as arising from non-interactive colonization to an Ŝ near the non-interactive Ŝ (and to a point above a more enduring inter-active Ŝ), followed by a slow decline in Ŝ (as population sizes and interaction increase) to an enduring Ŝ near that obtaining before defaunation. The effect on this scheme of increasing distance from source area is shown, and a simulation of the non-inter-active part of the scheme was performed which corroborates that part of the model. Explicit equations for the "expected" colonization, immigration, and island extinction curves are given for non-interactive colonization, and predictions are given about the general changes on these curves as interaction becomes significant.