AbstractThis research compares performance, reproducibility, and detection limits of ambient seawater analysis for trace metals using both borohydride reductive coprecipitation and an automated chelation column (seaFAST™ 2) preconcentration for matrix interferent elimination on total and dissolved grab samples in nearshore to marine waters, over a broad concentration range, prior to inductively coupled plasma mass spectrometry (ICP‐MS) injection. A move to an online preconcentration method both minimizes sample preparation, and eliminates correction errors when accounting for trace impurities in precipitated samples, induced via reagents. The reproducability of the online preconcentration method described, coupled with low blanks and method detection limits (MDLs), demonstrates the effectiveness of the automated procedure using ethylenediaminetriacetic and iminodiacetate acid chelation exchange resin and multianalyte determination by ICP‐MS for total and dissolved Ni, Cu, Zn, Cd, and Pb in marine water samples. Average CASS‐5 recoveries using the online preconcentration method (n = 9) were 109% ± 7%, 104% ± 5%, 103% ± 7%, 101% ± 3%, and 86% ± 8%, respectively. The MDLs obtained from the automated method for Ni, Cu, Zn, Cd, and Pb were 3.3, 1.8, 13.5, 4, and 10 times lower, respectively, than for the Borohydride method. There were statistically significant differences between the methods for CASS‐5 recoveries of Ni, Cu, Zn (p < 0.0001), and Pb (p = 0.0024). Comparison of methods gave high concordance (rC ≥ 0.90) between methods for total and dissolved Ni, Cu, Zn, and Pb, and total Cd.