Routine measurements of volatile organic compounds (VOCs) in air have shown that average concentrations are very much smaller than those used in laboratory experiments designed to study the effects of VOCs on plants. However, maximum hourly concentrations of some VOCs can be 100 times larger than the average, even in rural air. Experimental studies have rarely extended for longer than a few days, so there is little information on potential long-term effects of exposure to small concentrations. This review considers the available evidence for long-term effects, based on laboratory and field data. Previous reviews of the literature from Germany and the USA are cited, prior to an assessment of the effects of individual VOCs. Although hydrocarbons from vehicle exhausts have been implicated in the observed effects on roadside vegetation, the evidence suggests that it is the nitrogen oxides in the exhaust gases that are mostly responsible. There is evidence that aromatic hydrocarbons can be metabolised in plants, although the fate of the metabolites is not known. There is a large literature on the effects of ethylene, because of its role as a plant hormone. Effects have been reported in the field, in response to industrial emissions, and dose-response experiments over several weeks in laboratory studies have clearly identified the potential for effects at ambient concentrations. The main responses are morphological (e.g. epinasty), which may be reversible, and on the development of flowers and fruit. Effects on seed production may be positive or negative, depending on the exposure concentration. Chlorinated hydrocarbons have been identified as potentially harmful to vegetation, but only one long-term experiment has studied dose-response relationships. As for ethylene, the most sensitive indication of effect was on seed production, although long-term accumulation of trichloroacetic acid in tissue may also be a problem. There is little evidence of the direct effects of oxygenated hydrocarbons on plants. Plants are a significant emission source of short-chain alcohols, aldehydes and ketones. Peroxyacetyl nitrate (PAN) has a well-documented history as damaging to vegetation. There have been few long-term experimental studies despite the field evidence for damaging effects. Early studies in California have been followed by more recent data from east Asia, but there is still a dearth of information on the potential for effects of PAN and related peroxyacyl nitrates on vegetation typical of regions around tropical and sub-tropical cities where PAN pollution is increasingly important. The lack of long-term measurements, coupled with the available evidence that effects are not linearly related to 'dose' measured as the product of exposure concentration and time, means that the possibility of adverse effects of VOCs on vegetation cannot be safely rejected, particularly in urban and industrial areas. Although reproductive processes (flowering, seed production) appear to be most sensitive, there have been no experimental studies on subsequent seed viability and the consequences at the ecosystem level of changes to plant phenology. The potential for VOC metabolites to accumulate in plant tissue has been demonstrated, but any subsequent effects on herbivores and phytophagous insects have yet to be investigated.