The effects of the herbicide, DNOC, 4,6-dinitro-o-cresol, a model pollutant, have been studied by comparing the potential of soil microbial communities present in individual soil aggregates or in larger soil microcosms as samples of soil aggregates to mineralise the herbicide 2,4D. We have shown that 2-3 mm soil aggregates vary widely in their 2,4-D mineralisation potential and that ageing or exposure to DNOC considerably simplified the distribution patterns of this capacity. The main factors of variation have been quantified and classified using a quasi-likelihood method derived from the Generalised Linear Model approach. Besides DNOC concentration and duration of exposure, an additional 'rank' factor reflecting a desiccation gradient of the aggregates on the microliter plates was found to have statistical significance. We concluded that it should be possible to derive an experimental approach, designated as 'functional profiling', with potential use to detect soil chemical contamination. Curves of 2,4-D mineralisation in individual soil aggregates could be classified according to three different types of kinetics, which were assumed to reflect heterogeneous spatial distribution, differences in microbial community composition and varying efficiency of the microbial consortia involved in 2,4-D degradation. Exposure to DNOC considerably simplified the distribution patterns of the different types of kinetics with one type, showing slow rate and low cumulative mineralisation, becoming predominant as ageing, concentration and duration of DNOC exposure increased. We argue on the possible use of 'kinetic profiling' as a sensitive bioindicator of soil quality. By comparison, in soil microcosms, 2,4-D mineralisation showed an extra mineralisation potential of 64% over individual aggregates in the control soil and exposure to DNOC was followed by concentration and time-dependent recovery of the 2,4-D mineralisation potential. It is likely that 2 g size soil microcosms gather a larger number of biochemical capacities which could complement each other to increase the potential of soil to mineralise xenobiotic compounds. (c) 2004 Elsevier Ltd. All fights reserved.