Seawater samples from the Gulf of Alaska and estuarine water samples from the York River in Virginia near the Chesapeake Bay are directly analyzed for barium concentration using graphite furnace atomic emission spectrometry. The accuracy of the method is evaluated by a correlation of results with data from analyses using line source and wavelength-modulation continuum-source atomic absorption spectrometry and graphite furnace atomization, as well as the analysis of a standard reference water sample. Interferences in the analysis and techniques used to overcome them are discussed.T h e distribution of barium in seawater is of great significance t o t h e study of the marine environment. T h e concentration of barium in the oceans has been proposed as a potential indicator of marine drilling contamination (1 ), as an indicator of biologically productive oceanic areas ( 2 ) , and as a chemical analogue of radium, which may be useful for the study of deep water circulation processes (3). T h e barium concentration in seawater is regulated by t h e sulfate concentration, based on t h e solubility of BaSO,, and has been reported to vary in concentration between 3 t o 105 pg/L depending on t h e sampling location and depth (1).Methods for quantitation of barium in seawater are therefore of great concern to marine scientists. Analysis for barium in seawater has been performed by neutron activation (4-61, mass spectrometry (1-3, 7 , 8), and flame atomic spectrometry with concentration techniques (9, 10). However. all of these techniques are either time-consuming or expensive, and the need still exists for a simple, fast, and accurate method for barium quantitation.T h e use of the graphite furnace has extended the application of atomic absorption spectrometry (AAS) to elemental concentrations several orders of magnitude lower than previously possible using flame techniques. However, the low concentrations of barium found in some locations in the ocean are near the limit of detection of even the furnace atomic absorption technique, and t h e analysis has been further complicated by the absence, until recently ( 1 2 , 12), of adequate background correction procedures for use in the visible region of t h e spectrum. Scatter and/or molecular absorption due to the seawater matrix may thus introduce a significant error in barium determinations by furnace atomic absorption.T h e measurement of emission from a graphite tube furnace has been shown to be analytically useful by several authors (13,14). Barium is one of the elements for which graphite furnace atomic emission spectrometry (GFAES) is extremely sensitive (15,16). T h e emission detection limit is a t least an order of magnitude lower than t h a t obtained by the typical graphite furnace atomic absorption spectrometry (GFAAS) analysis.Present address, Seawater samples collected from the Gulf of Alaska and estuarine water from the York River entering the Chesapeake Bay in Virginia are used to evaluate the technique of furnace atomic emission for the quantitation of barium...