An open-air fumigation system was used to expose a winter barley crop (Hordeum vulgare L.) to controlled concentrations of SO2 from November 1982 to July 1983. Monitoring data from a rural site were used as a basis for simulating mean concentrations of 0.023, 0.038 and 0.058 μl 1−1 SO2 at a location where background concentrations were 0.01μl 1−1 SO2. There was no evidence of a reduction or stimulation of crop growth over the winter period. During the spring, the crop in SO2 fumigated plots had a higher green leaf area and dry weight than in control plots which may have resulted from a reduction in the occurrence of the fungal pathogen brown rust (Puccinia Hordei Otth.). After car emergence a delay in senescence of leaves was observed in fumigated plots which gave the plants an extended period for assimilation during grain filling. There was no evidence that the yield of barley was reduced within the range of ambient SO2 concentration in the UK (annual mean 0.03μ1 l−1SO2). However, at 0.058μl 1−1 SO2, reductions in yield were observed. Treatment with 0.023μl 1−1 SO2 produced increased grain yield.