To estimate potential O-3-induced crop losses during the last nine years (1984-1992) in the state of Hessen, Germany, we calculated exposure statistics from statewide records of surface O-3 concentrations and used these data as input for published O-3 exposure-response functions of several major crops. Estimated potential yield losses for a ''background level'' of O-3, which roughly reflects pre-industrial conditions, were 0.1% (barley and an O-3-tolerant winter wheat cultivar in 1984) in comparison to 32.6% (an O-3-sensitive spring wheat cultivar in 1992) under present-day conditions, with an increasing trend in response since 1984. However, available O-3 exposure-response functions were derived from experiments with controlled O-3, concentrations, but generally under present-day CO2 supply. Since current CO2 concentrations exceed pre-industrial levels by about 26%, a potential CO2 fertilization effect has to be considered, when assessing crop losses to ozone relative to pre-industrial atmospheric conditions. For two spring wheat cultivars a data set from different experiments was available, allowing us to account for CO2 fertilization and O-3 effects simultaneously assuming a linear-additive crop response to CO2 and O-3. The model equations were subsequently used to describe the time course of relative yield performance of these wheat cultivars since 1900 as a function of different scenarios of CO2 and O-3 in the atmosphere. Based on these results, the implications for the design of current experimental studies on crop responses to CO2 and O-3 are discussed.