Soil nematode feeding groups are a long-established trophic categorisation largely based on morphology and are used in ecological indices to monitor and analyse the biological state of soils. Stable isotope ratio analysis (13C/12C and 15N/14N, expressed as δ13C and δ15N) has provided verification of, and novel insights into, the feeding ecology of soil animals such as earthworms and mites. However, isotopic studies of soil nematodes have been limited to date as conventional stable isotope ratio analysis needs impractically large numbers of nematodes (up to 1,000) to achieve required minimum sample weights (typically >100 µg C and N). Here, micro-sample near-conventional elemental analysis–isotopic ratio mass spectrometry (μEA–IRMS) of C and N using microgram samples (typically 20 µg dry weight), was employed to compare the trophic position of selected soil nematode taxa from four feeding groups: predators (Anatonchus and Mononchus), bacterial feeders (Plectus and Rhabditis), omnivores (Aporcelaimidae and Qudsianematidae) and plant feeder (Rotylenchus). Free-living nematodes were collected from conventionally and organically managed arable soils. As few as 15 nematodes, for omnivores and predators, were sufficient to reach the 20 µg dry weight target. There was no significant difference in δ15N (p = 0.290) or δ13C (p = 0.706) between conventional and organic agronomic treatments but, within treatments, there was a significant difference in N and C stable isotope ratios between the plant feeder, Rotylenchus (δ15N = 1.08 to 3.22 mUr‰, δ13C = –29.58 to –27.87 mUr) and all other groups. There was an average difference of 9.62 mUr in δ15N between the plant feeder and the predator group (δ15N = 9.89 to 12.79 mUr, δ13C = –27.04 to –25.51 mUr). Isotopic niche widths were calculated as Bayesian derived standard ellipse areas and were smallest for the plant feeder (1.37 mUr2) and the predators (1.73 mUr2), but largest for omnivores (3.83 mUr2). These data may reflect more preferential feeding by the plant feeder and predators, as assumed by classical morphology-based feeding groups, and indicate that omnivory may be more widespread across detritivore groups i.e. bacterial feeders (3.81 mUr2). Trophic information for soil nematodes derived from stable isotope analysis, scaled as finely as species level in some cases, will complement existing indices for soil biological assessment and monitoring, and can potentially be used to identify new trophic interactions in soils. The isotopic technique used here, to compare nematode feeding group members largely confirm their trophic relations based on morphological studies.