Stefanie von Felten, Nina Buchmann, and Michael Scherer-Lorenzen Harrison et al. (2007) reported on an interesting N labeling study. Under field conditions, they assessed whether coexisting plant species of temperate grasslands show preferences for different chemical forms of nitrogen (N), including ammonium nitrate (inorganic N) and three amino acids of varying complexity (organic N). The authors found that all plant species were able to take up the full range of amino acids offered to them, as shown by N and C enrichment in plant tissues. However, plants all preferred inorganic over organic N, indicated by higher N enrichments after ammonium nitrate compared to organic N labeling. We do not object to the general interpretation of the results and the authors’ main conclusions. Yet, we would like to comment on the plant uptake of intact amino acids. When testing for significant relationships between excess C and N of plants to infer direct uptake of amino acids (Nasholm et al. 1998), Harrison et al. (2007) should have accounted for the different C:N ratios of the amino acids used. The amino acid tracers were U-C2N-glycine, U-C3N-serine, and U-C9N-phenylalanine (all N 98% and C 98%), and their ratios of C:N atoms are 2:1, 3:1, and 9:1 respectively. While the authors point out that these differences in available C may affect the preferences of plants and microbes, they omitted to consider the methodological consequences. One common problem (see e.g., Jones et al. 2005) when using dual-labeled amino acids to study organic N uptake by plants is detecting the C label in plants. Due to the high C:N ratio of plants and the high abundance of C (;1.08 atom % in C3 plants), the dilution of C is usually 60–150 times higher than that of N (Nasholm and Persson 2001). Finding a significant relationship between excess C and N requires separating the shift in C resulting from direct amino acid uptake from natural variation and analytical error. However, this is often not possible, due to rather low concentrations of tracer C. As a solution, Nasholm and Persson (2001) suggested to concentrate the labeled fraction of the plant material studied, by extracting the soluble fraction containing the label. For assessing the uptake of intact amino acids using the dual-labeling approach, the critical step is to assure that there is a theoretical possibility of detecting this uptake. From the measured values of dN (after labeling with N) the theoretical shift in dC corresponding to 100% intact uptake can be calculated (Nasholm and Persson 2001). Thereby it can be determined whether this shift is distinguishable from ‘‘noise.’’ Given the high amount of C in phenylalanine, it is not surprising that Harrison et al. (2007) found a significant relationship between excess C and N across all species for this amino acid, but not for glycine and serine. In their paper, Fig. 2A shows that shoot N enrichment over all plant species was highest for glycine and lowest for phenylalanine (among organic N forms), while shoot C enrichment was similar for all amino acids (Fig. 2C). This almost opposite pattern for C and N enrichment also applies for single species (Fig. 1), roots (Fig. 3), and microbes (Fig. 4). In the latter, C enrichment was actually highest when labeled with phenylalanine, and lowest in the case of glycine. We think that these results are due to the different C:N ratios of the three amino acids rather than indicating higher uptake of phenylalanine compared to glycine and serine, which is particularly unlikely given that phenylalanine is the largest and most complex amino acid tested. However, without significant relationships between excess C and N in plant tissues, the proportion of amino acids taken up as intact molecule cannot be estimated for glycine and serine. Moreover, although no data on amino acid concentration in the soil solution are shown, it is likely that phenylalanine is the least abundant of the three amino acids, and glycine the most abundant. Thus, the dilution of the added N tracer (equal for all N forms) by the natural abundance pool was probably least for phenylalanine and strongest for glycine, again leading to an overestimation of phenylalanine uptake when assessed by N labeling. We fully agree with Harrison et al. (2007), that a rigorous test to detect organic N uptake by plants requires compound specific isotope analysis (a combination of gas chromatography with isotope ratio mass spectrometry; see e.g., Persson and Nasholm 2001). But clearly, the results of Harrison et al. (2007) demonstrate that the use of the Nasholm et al. (1998) method to infer Manuscript received 21 June 2007; accepted 27 August 2007. Corresponding Editor: P. M. Groffman. 1 Institute of Plant Sciences, ETH Zurich, Zurich CH-8092 Switzerland. 2 Institute of Environmental Sciences, University of Zurich, Zurich CH-8057 Switzerland. E-mail: stefanie.vonfelten@ipw.agrl.ethz.ch