Environmental degradation of organic micropollutants is difficult to monitor due to their diffuse and ubiquitous input. Current approaches-concentration measurements over time, or daughter-to-parent compound ratios-may fall short, because they do not consider dilution, compound-specific sorption characteristics or alternative degradation pathways. Compound-specific isotope analysis (CSIA) offers an alternative approach based on evidence from isotope values. Until now, however, the relatively high limits for precise isotope analysis by gas chromatography-isotope ratio mass spectrometry (GC-IRMS) have impeded CSIA of sub-microgram-per-liter scale micropollutant concentrations in field samples. This study presents the first measurements of C and N isotope ratios of the herbicide atrazine and its metabolite desethylatrazine at concentrations of 100 to 1,000 ng/L in natural groundwater samples. Solid-phase extraction and preparative HPLC were tested and validated for preconcentration and cleanup of groundwater samples of up to 10 L without bias by isotope effects. Matrix interferences after solid-phase extraction could be greatly reduced by a preparative HPLC cleanup step prior to GC-IRMS analysis. Sensitivity was increased by a factor of 6 to 8 by changing the injection method from large-volume to cold-on-column injection on the GC-IRMS system. Carbon and nitrogen isotope values of field samples showed no obvious correlation with concentrations or desethylatrazine-to-atrazine ratios. Contrary to expectations, however, δ (13) C values of desethylatrazine were consistently less negative than those of atrazine from the same sites. Potentially, this line of evidence may contain information about further desethylatrazine degradation. In such a case, the common practice of using desethylatrazine-to-atrazine ratios would underestimate natural atrazine degradation.