A primary concern in applying existing O3-effects data on crop production is the relatively unknown influence of soil moisture which may modify plant response to O3. One of the deficiencies in performing field experiments to test the influence of soil moisture on crop response to O3 has been the lack of control of soil moisture conditions in open-top chamber plots. In 1986 a field experiment was performed with soybean to test the influence of moderate water stress on the yield response to O3. Open-top field chambers were used to expose plants to a range of O3 concentrations, and rain exclusion caps were used on individual chambers to help regulate soil moisture levels. There were three soil moisture treatments (well-watered (WW), water-stressed (WS), and well-watered with permanent rain exclusion caps that were in place from 35 days after planting until physiological maturity (WW-C)). In the WW and WS treatments, the rain caps were put in place only during an exceptionally wet period from mid-August to mid-September. The WW and WW-C treatments had approximately the same yield and a similar response to O3, indicating that the presence of the caps for most of the growing season had little effect on growth or sensitivity to O3. The WS plots yielded 15 to 18% less that the WW and WW-C plots, but water stress did not change the relative response to O3 (i.e., no significant O3 x water interaction). Based on a Weibull dose-response model, O3 reduced the yield of 'Young' soybean 13% at a concentration of 0.05 microL/L (12 h/d seasonal mean) compared to a hypothetical background of 0.02 microL/L.