Development and use of ozone (O-3)-resistant crop cultivars are key measures to avoid agricultural yield reduction in a high O-3 environment. However, little is known about the impacts of breeding methods and breeding sites on the development of O-3 tolerance in winter wheat cultivars. To explore such impacts, 20 Chinese winter wheat cultivars bred using four breeding methods (viz. introduction, reselection, conventional breeding and hybridization) at four breeding sites having different levels of O-3 exposures, were exposed to charcoal-filtered (CF) air or high O-3 (82 ppb, 7 h d(-1)) for 21 days. O-3 tolerance of cultivars was assessed by the relative levels of visible injury, growth, gas exchange, dark respiration, antioxidative activities and oxidative modification of proteins and cellular membranes. We found that conventional breeding and hybridization demonstrated higher potential capacity for O-3 tolerance as indicated by a higher level of ascorbate and peroxidase activity in cultivars exposed to CIF air. Despite the highest potential capacity for O-3 tolerance, hybridization displayed the lowest O-3 tolerance as represented by antioxidative activities, oxidative stress, photosynthesis and growth. The causes of higher O-3 sensitivity in hybrid cultivars included lower O-3 exclusion by stomatal closure, higher reduction in antioxidative activities, higher O-3-induced modification of proteins and cellular membranes, lower level of repair of O-3-induced cellular damage and higher loss of assimilation rate as well as growth in O-3 relative to control plants. Cultivars bred at breeding sites experiencing higher ambient O-3 exposures demonstrated higher potential capacity for O-3 tolerance. The observed O-3 tolerance in cultivars; bred by different breeding methods was uncorrelated to ambient O-3 levels in breeding sites as well as to the potential O-3 tolerance capacity as observed in CF air. Results from our experiment, therefore, clearly indicated that potential O-3 sensitivity would have little use in predicting actual O-3 tolerance of winter wheat cultivars. Our findings also suggested that sensitivity to O-3 in winter wheat cultivars was related to breeding methods, but not to O-3 concentrations in breeding sites. (C) 2009 Elsevier B.V. All rights reserved.