A brief account is given of the ecological significance of quantitatively important secondary plant compounds, mainly those of a phenolic nature, in herbivory and decomposition. Phenolic compounds accumulate to a greater extent in slow-growing species than in fast-growing ones, particularly when soil conditions (nutrients, water) restrict growth. Two hypotheses to explain the increased concentration of phenolics when soil conditions are unfavorable are presented. The first hypothesis (the 'carbon supply model of secondary plant metabolism') considers the increased levels of non-structural carbohydrates as the major trigger. The second hypothesis (the 'amino acid diversion model of secondary plant metabolism') states that increased accumulation of phenolics stems from a decreased use of a common precursor (phenylalanine or tyrosine) for protein synthesis. Current experimental evidence, though still fairly limited, supports the second hypothesis, but further testing is required before the first model can be rejected. So far, there is very little evidence for a direct effect of atmospheric CO2 on the concentration of secondary compounds in higher plants. However, there are likely to be indirect effects, due to a stronger limitation by the nitrogen supply in plants whose growth has been promoted by atmospheric CO2. It is concluded that it is very likely that phenolic compounds accumulate to a greater extent in plants exposed to elevated CO2, due to a greater limitation of nutrients, rather than as a direct effect of elevated CO2.