The ubiquity and abundance of nematodes in soils and aquatic sediments interest ecologists seeking models for fluxes of matter and energy. However, the complete spectrum of roles and the quantitative importance of nematodes in soils and aquatic sediments remain poorly documented. Feeding type classifications of nematodes usually rely strongly on buccal morphology, but there are many unresolved questions on food sources and feeding rates. Here we comment on the much used feeding type classification of Yeates et al (1993a), which, pending minor modifications, we still consider valid. However, we argue against the use of the separate category, omnivores. We also advocate the use of this terrestrial nematode-based scheme in aquatic nematology, including marine nematodes. A potentially major shortcoming of feeding type classifications is that they attribute guild- rather than species-specific functions to nematodes. This conflicts with evidence of high feeding selectivity and flexibility. Nematode species diversity at small spatial scales thus provides an intriguing system for studies on the functional implications of diversity and of species redundancy. Finally, to relate laboratory-derived feeding rates to population growth rates under natural conditions, we address the effects of external influences on nematode activity in soils, and stress the need for a better understanding of nematode functional responses.