To evaluate nitrogen (N) saturation in xeric environments, we measured hydrologic N losses, soil N pools, and microbial processes, and developed an N-budget for a chaparral catchment (Sierra Nevada, California) exposed to atmospheric N inputs of approximately 8.5 kg N ha -1 y -1. Dual-isotopic techniques were used to trace the sources and processes controlling nitrate (NO 3 -) losses. The majority of N inputs occurred as ammonium. At the onset of the wet season (November to April), we observed elevated streamwater NO 3 -concentrations (up to 520 lmol l -1 ), concomitant with the period of highest gaseous N-loss (up to 500 ng N m -2 s -1 ) and suggesting N-saturation. Stream NO 3 -d
15N and d
18O and soil N measurements indicate that nitrification controlled NO 3 -losses and that less than 1% of the loss was of atmospheric origin. During the late wet season, stream NO 3 -concentrations decreased (to <2 lmol l -1 ) as did gaseous N emissions, together suggesting conditions no longer indicative of N-saturation. We propose that chaparral catchments are temporarily N-saturated at £ 8.5 kg N ha -1 y -1 , but that N-saturation may be difficult to reach in ecosystems that inherently leak N, thereby confounding the application of N-saturation indicators and annual N-budgets. We propose that activation of N sinks during the typically rainy winter growing season should be incorporated into the assessment of ecosystem response to N deposition. Specifically, the N-saturation status of chaparral may be better assessed by how rapidly catchments transition from N-loss to N-retention.