Atmospheric inputs of iron and nitrogen have been hypothesized to play an important role in the chemical and biological dynamics of open ocean ecosystems (Menzel and Spaeth, 1962; Duce, 1986). This hypothesis has stimulated considerable discussion and controversy (see Martin, 1991, this issue; Miller et al., 1991, this issue; Morel et al., 1991, this issue). Much of this discussion has focused on the relative importance of atmospheric nutrient input rates over large time and space scales, the physiology and chemistry of iron nutrition, the elemental composition of phytoplankton (e.g., Fe/C), and the interpretation of assays for iron limitation of phytoplankton growth at a particular time and place. An interdisciplinary analysis has led to the realization that biological responses of oceanic ecosystems are likely to be dependent not only on biological requirements and the large-scale variability in annual-average atmospheric inputs but also on the temporal and spatial variability of these inputs, on the chemical and biological processes controlling the fate of those inputs, and on the biological, chemical, and physical interactions controlling the dynamics of open-ocean ecosystems receiving those inputs. The results of this interdisciplinary analysis are summarized below.