Abstract Isoprene is synthesized and emitted in large amounts by a number of plant species, especially oak (Quercus sp.) and aspen (Populussp.) trees. It has been suggested that isoprene improves thermotolerance by helping photosynthesis cope with high temperature. However, the evidence for the thermotolerance hypothesis is indirect and one of three methods used to support this hypothesis has recently been called into question. More direct evidence required new methods of controlling endogenous isoprene. An inhibitor of the deoxyxylulose 5-phosphate pathway, the alternative pathway to the mevalonic acid pathway and the pathway by which isoprene is made, is now available. Fosmidomycin eliminates isoprene emission without affecting photosynthesis for several hours after feeding to detached leaves. Photosynthesis of fosmidomycin-fed leaves recovered less following a 2-min high-temperature treatment at 46°C than did photosynthesis of leaves fed water or fosmidomycin-fed leaves in air supplemented with isoprene. Photosynthesis of Phaseolus vulgaris leaves, which do not make isoprene, exhibited increased thermotolerance when isoprene was supplied in the airstream flowing over the leaf. Other short-chain alkenes also improved thermotolerance, whereas alkanes reduced thermotolerance. It is concluded that thermotolerance of photosynthesis is a substantial benefit to plants that make isoprene and that this benefit explains why plants make isoprene. The effect may be a general hydrocarbon effect and related to the double bonds in the isoprene molecule.