Suppositions that the trichloroacetic acid (TCA, CCl3C(O)OH) found in nature was a consequence solely of the use of chlorinated hydrocarbon solvents prompted this critical review of the literature on its environmental fluxes and occurrences. TCA is widely distributed in forest soils (where it was rarely used as an herbicide) and measurements suggest a soil flux of 160 000 tonnes yr(-1) in European forests alone. TCA is also produced during oxidative water treatment and the global flux could amount to 55 000 tonnes yr(-1) (from pulp and paper manufacture, potable water and cooling water treatments). By contrast, the yields of TCA from chlorinated hydrocarbon solvents are small: from tetrachloroethene 13 600 tonnes yr(-1) and from 1,1,1-trichloroethane 4300 tonnes yr(-1) on a global basis, at the atmospheric burdens and removal rates typical of the late 1990s. TCA is ubiquitous in rainwater and snow. Its concentrations are highly variable and the variations cannot be connected with location or date. However, there is no significant difference between the concentrations found in Chile and in eastern Canada (by the same analysts), or between Malawi and western Canada, or between Antarctica and Switzerland, nor any significant difference globally between the concentrations in cloud, rain and snow (although local enhancement in fog water has been shown). TCA is present in old ice and firn. At the deepest levels, the firn was deposited early in the 19th century, well before the possibility of contamination by industrial production of reactive chlorine, implying a non-industrial background. This proposition is supported by plume measurements from pulp mills in Finland. TCA is ubiquitous in soils; concentrations are very variable but there are some indications that soils under coniferous trees contain higher amounts. The concentrations of TCA found in plant tissue are region-specific and may also be plant-specific, to the extent that conifers seem to contain more than other species. TCA is removed from the environment naturally. There is abundant evidence that soil microorganisms dehalogenate TCA and it is lost from within spruce needles with a half-life of 10 days. There is also recent evidence of an abiotic aqueous decarboxylation mechanism with a half-life of 22 days. The supposedly widespread effects of TCA in conifer needles are not shown in controlled experiments. At concentrations in the needles of Scots pine similar to those observed in needles in forest trees, changes consequent on TCA treatment of field laboratory specimens were almost all insignificant.