We investigated the distribution and abundance of juvenile flyingfish in the eastern Caribbean by mghtlighting and dipnetting at 20 stations from April 10 to May 6, 1988. Of the 2211 fish collected, 73.5 % were flym@sh. Abundance of juvenile flying-fish varied significantly across the survey area, but was not correlated with any of the surface water characteristics measured (salinity, temperature, PO4-P, NO3-NO2-N), nor with flotsam abundance or moon phase. Parexocoetus brachypterus (49.5 %), Exocoetus volitans (41.0%) and Hirundichthys affuns (8.0%) accounted for 98.5 % of the flyingfish catch. The length-frequency distribution of P. brachypterus was trimodal, the largest size group being adult. The length-frequency distributions of E. volitans and H. affinis were bimodal. In both cases, the largest size group was adult, but adults of E. volitans were rare, accounting for < 3 % of the E. volitans catch. Abundance of all size groups of all 3 species varied significantly across the survey area, and different life history stages within a species had different geographical distributions. For example, adults of E. vohtans were virtually absent from the survey area, but larvae and juveniles were common. Variation in abundance of juveniles and adults of H. affinis was not correlated across the survey area, and the same was true for small juveniles and adults of P. brachypterus. The integrity of sizespecific geographical distributions is supported by the fact that different survey methods typically detected the same geographical distributions for similar life history stages. Abundance of P. brachypterus larvae in neuston tows was correlated with abundance of the smaller juvenile size group of P. brachypterus sampled by nightlighting. Abundance of E. volitans larvae in neuston tows was correlated with abundance of E. volitans juveniles taken by nightlighting, and abundance of H. affinis adults detected by visual survey was correlated with their abundance in mghtlightmg samples.