The analytical estimation of amorphous silica in selected Atlantic and Antarctic Ocean sediments, the U.S.G.S. standard marine mud (MAG-1), A.A.P.G. clays, and samples from cultures of a marine diatom, Hemidiscus, has been examined. Quantitative recovery of sedimentary amorphous silica was achieved by reacting 2 M Na 2 CO 3 with a sample for 4 hours at 90 degrees -100 degrees C; where necessary, aluminum analysis is used to correct for the extraction of non-amorphous silica. Oceanic sediments having an amorphous SiO 2 /clay ratio of 1.0 or more can be analyzed by a single extraction with 2 M Na 2 CO 3 , without correction for non-amorphous additions. Marine sediments having an amorphous SiO 2 /clay ratio of from 1.0 to 0.25 can be corrected for inputs of clay-derived silica using an aluminum determination and an empirical correction factor. Sediments with amorphous SiO 2 /clay ratios of less than 0.25, samples containing non-silica-bearing aluminous materials such as gibbsite, or materials having low absolute levels of amorphous silica require successive leaches to accurately correct for silica inputs from non-amorphous sources. Our values for amorphous silica-rich circum-Antarctic sediments are equal to or greater than literature values, whereas our values for a set of amorphous silica-poor sediments from a transect of the North Atlantic at 11 degrees N, after appropriate correction for silica released from clays, are significantly lower than previous estimates from the same region.