Mass spectrometer analysis of highly enriched 15N samples underestimated the 15N enrichment. Conventional 15N analysis utilizes a ratiometer procedure in which the ratio of 29N2 to 28N2 ions is determined. Samples containing high enrichments had a low quantity of 28N2 ions and it appeared background contamination of 28N2 produced erroneous results with this procedure. Therefore, results obtained with the ratiometer procedure were compared to those obtained with a scanning procedure which measured the 28N2, 29N2 and 30N2 peak heights. The 15N enrichment was determined by the scanning procedure by using all three N2 peaks (28/29/30 method), the 28N2 and 29N2 peaks (28/29 method), and 29N2 and 30N2 peaks (29/30 method). The four methods were compared over a range of enrichments from 0.366 to 99 atom % 15N in samples containing 0.5, 1.0 and 1.5 mg N. Samples with enrichments less than 10 atom % 15N were best estimated by the ratiometer method. Enrichments greater or equal to 10 atom % 15N were most accurately determined by the 29/30 method. Background contamination during sample preparation produced 28N2 peaks which had a negligible effect on samples containing low enrichments, but was a source of error in highly enriched samples. Since natural levels of 29N2 and 30N2 in air are very low, the 29/30 method was found to closely estimate the expected 15N enrichment in standard samples greater than 10 atom % and had the lowest standard deviation.