A technique has been developed to quantify ultratrace 231Pa (50-2000 ag; 1 ag = 10-18 g) concentrations in seawater using isotope-dilution thermal ionization mass spectrometry (TIMS). The method is a modification of a process developed by Pickett et al. (Pickett, D. A.; Murrell, M. T.; Williams, R. W. Anal. Chem. 1994, 66, 1044-1049) and extends the technique to very low levels of protactinium. The procedural blank is 16 ± 15 ag (2σ), and the ionization efficiency (ions generated/atom loaded) approaches 0.5%. Measurement time is <1 h. The amount of 231Pa needed to produce 231Pa data with an uncertainty of ±4-12% is 100-1000 ag (∼3 × 105 to 3 × 106 atoms). Replicate measurements made on known standards and seawater samples demonstrate that the analytical precision approximates that expected from counting statistics and that, based on detection limits of 38 and 49 ag, protactinium can be detected in a minimum sample size of surface seawater of ∼2 L for suspended particulate matter and <0.1 L for filtered (<0.4 μm) seawater, respectively. The concentration of 231Pa (tens of attograms per liter) can be determined with an uncertainty of ±5-10% (2σ) for suspended particulate matter filtered from 5 to 10 L of seawater. For the dissolved fraction, 0.5-1 L of seawater yields 231Pa measurements with a precision of 1-10%. Sample size requirements are orders of magnitude less than traditional decay-counting techniques and significantly less than previously reported ICP-MS techniques. Our technique can also be applied to other environmental samples, including cave waters, rivers, and igneous rocks.