Both sensitive and tolerant clones of aspen (Populus tremuloides) were exposed to ozone using four different exposure regimes under controlled environmental conditions. Based on data on ambient ozone from 10 cities in the USA, three treatments of 4-wk exposure to the same SUM06 (an accumulation of hourly O3 concentrations greater than 0.06 ml l−1) were constructed. The regimes allowed us to investigate: (a) the importance of long (3 wk, treatment 1) versus short (1 wk, treatment 2) duration of regimes with high peaks; (b) the effect of treatments with variable peak occurrence (treatments 1 and 2) versus uniform peak occurrence (treatment 3) during the exposure period. Nonfumigated control plants were maintained at ozone concentrations <10 nl l−1. Bifacial black necrosis, a typical symptom of ozone injury on aspen leaves, occurred on both clones after 2 wk exposure. Up to 60% of the leaves on the sensitive clone were injured, with an average of 6% of total leaf area injured. In the tolerant clone only 10% of the leaves were injured, with less than 1% of the total leaf area symptomatic. The severity of injury was consistently greatest in treatment 2, followed by treatments 1 and 3, respectively. The interval between peak exposures was less important than the occurrence of peaks versus a stable maximum concentration. Premature leaf abscission occurred in the sensitive clone. Measures of gas exchange demonstrated reduced photosynthesis under ozone fumigation, but exposure regime was not a significant factor. Concentrations of two antioxidants, ascorbic acid and glutathione, were almost always greater in the resistant than in the sensitive clone, but the differences were not statistically significant. The levels of these antioxidants in aspen leaves did not change with ozone fumigation or leaf age.