Determination of the isotopic signature of dissolved organic nitrogen (DON) is important to assess its dynamics in terrestrial ecosystems. Analysis of (15)N-DON, however, has been hindered by the lack of simple, reliable, and established methods. We evaluate three off-line techniques for measuring the (15)N signature of DON in the presence of inorganic N using a persulfate digestion followed by microdiffusion. The (15)N-DON signature is calculated from the difference between total dissolved (15)N ((15)N-TDN) and inorganic (15)N. We quantified the (15)N recovery and signature of DON, NH(4)(+), and NO(3)(-) in a series of inorganic N/DON mixtures (with a TDN concentration of 10 mg N L(-1)) for three lab protocols. Phenylalanine was used as a model compound for DON. The best lab protocol determined the concentration of inorganic N and TDN prior to diffusion using improved spectrophotometric techniques. An accuracy of 88% for (15)N-DON should be routinely possible; coefficient of variation was <2.9%. Hence, reliable (15)N-DON values are obtained over an DON concentration range of 2.3-10 mg L(-1). High levels of DON could influence the accuracy of (15)N-NO(3)(-) mainly at DON:NO(3)(-) ratios above 0.4. Evaluation of alternative NO(3)(-) measurements is still necessary. Our method is applicable for soil solution samples and soil extracts and has no risk of cross-contamination. Potential applications are large, in particular for (15)N tracer studies, and will increase our insight in DON behavior in soils.