The relative abundances of planktonic Foraminifera, oxygen-isotope values of benthic Foraminifera, and the sedimentology of 156 samples from Deep Sea Drilling Project–International Phase of Ocean Drilling Site 548 in the northeast Atlantic Ocean, have been analyzed for the time span from about 3.5 to 2.0 m.y. B.P. The planktonic assemblage at Site 548 in the late Pliocene consisted mainly of subpolar to temperate species; warm-water species were rare. The planktonic record shows that waters in the northeast Atlantic were warm-temperate before 3.4 m.y. ago and gradually changed to subpolar, through a series of steps, by 2.0 m.y. B.P. An initial cooling event occurred between about 3.4 and 3.1 m.y. B.P. The beginning of this cooling event preceded the global benthic isotopic shift at 3.2 to 3.1 m.y. B.P., but the isotopic shift correlates to a marked change in the planktonic assemblage, suggesting that the cause of the shift also affected northeast Atlantic surface circulation. Following the 3.4 to 3.1 m.y. B.P. cooling event, the assemblages at Site 548 indicate a return of temperate waters. At this time, however, the assemblages also became quite variable with a clear cyclicity. From about 2.6 m.y. B.P. onward, there is a progressive cooling recorded at Site 548, leading to marked glaciations beginning at about 2.4 m.y. B.P. About 2.0 m.y. B.P. the planktonic assemblages were dominated by cold-water species and became quite stable, contrasting markedly with the immediately preceding interval of glaciation and extreme variability. The cyclicity seen in the faunal data after about 3.1 m.y. B.P. is principally of one period, with a cycle length closest to that of the Earth's obliquity and its influence on global insolation. The effect of obliquity on the Earth's insolation is most strongly felt at high latitudes, and so the cyclicity at Site 548 indicates that conditions in the Arctic have exerted an influence on the northeast Atlantic since at least 3 m.y. ago. Calculation of global insolation fluctuations for the late Pliocene shows that the major changes in the planktonic record at Site 548 cannot be explained by insolation variation, except for the period of stability at 2.0 m.y. B.P., which corresponds to a time of stable insolation. The planktonic data, however, do contain cyclicity that can be related to the periodicity of the Earth's orbital elements.