Photosynthesis and light O(2)-uptake of the aerial portion of the CAM plant Ananas comosus (L.) merr. were studied by CO(2) and O(2) gas exchange measurements. The amount of CO(2) which was fixed during a complete day-night cycle was equal to the amount of total net O(2) evolved. This finding justifies the assumption that in each time interval of the light period, the difference between the rates of net O(2)-evolution and of net light atmospheric CO(2)-uptake give the rates of malate-decarboxylation-dependent CO(2) assimilation. Based upon this hypothesis, the following photosynthetic characteristics were observed: (a) From the onset of the light to midphase IV of CAM, the photosynthetic quotient (net O(2) evolved/net CO(2) fixed) was higher than 1. This indicates that malate-decarboxylation supplied CO(2) for the photosynthetic carbon reduction cycle during this period. (b) In phase III and early phase IV, the rate of CO(2) assimilation deduced from net O(2)-evolution was 3 times higher than the maximum rate of atmospheric CO(2)-fixation during phase IV. A conceivable explanation for this stimulation of photosynthesis is that the intracellular CO(2)-concentration was high because of malate decarboxylation. (c) During the final hours of the light period, the photosynthetic quotient decreased below 1. This may be the result of CO(2)-fixation by phosphoenolpyruvate-carboxylase activity and malate accumulation. Based upon this hypothesis, the gas exchange data indicates that at least 50% of the CO(2) fixed during the last hour of the light period was stored as malate. Light O(2)-uptake determined with (18)O(2) showed two remarkable characteristics: from the onset of the light until midphase IV the rate of O(2)-uptake increased progressively; during the following part of the light period, the rate of O(2)-uptake was 3.5 times higher than the maximum rate of CO(2)-uptake. When malate decarboxylation was reduced or suppressed after a night in a CO(2)-free atmosphere or in continuous illumination, the rate of O(2)-uptake was higher than in the control. This supports the hypothesis that the low rate of O(2)-uptake in the first part of the light period is due to the inhibition of photorespiration by increased intracellular CO(2) concentration because of malate decarboxylation. In view of the law of gas diffusion and the kinetic properties of the ribulose-1,5-bisphosphate carboxylase/oxygenase, O(2) and CO(2) gas exchange suggest that at the end of the light period the intracellular CO(2) concentration was very low. We propose that the high ratio of O(2)-uptake/CO(2)-fixation is principally caused by the stimulation of photorespiration during this period.