Thylakoid particle preparations of the filamentous cyanobacterium Oscillatoria chalybea were prepared from cultures that were grown on nitrate, ammonium sulfate or arginine as the sole nitrogen source in the medium. Moreover, cultures that had no exogenous nitrogen source in the medium, hence used air as nitrogen source, were also grown. Oxygen gas exchange of these preparations was analyzed by means of mass spectrometry in an assay system that contained in the liquid phase only H216O-containing buffer (hence no H218O label) and air as gas phasein which part of the 16O2 had been substituted by 18O2. Oxygen uptake in the dark was very low in all preparations but reacted in a subtle way already to very dim (less than 10 lx) illumination. In preparations of nitrate- and ammonium sulfate-grown cultures the dark-uptake rate was enhanced in dim light and appeared further enhanced in the dark period following this illumination. In preparations of arginine-grown cells or cells without nitrogen source in the medium dim light inhibited dark oxygen uptake. In weak light (80 lx) none of the preparations showed an O2-uptake phenomenon but rather an 02-evolution phenomenon. Both mass signals (mass 32 and mass 36) indicated O2 evolution. No mixed label 16O18O was found indicating that 18O2 evolved cannot come from water produced by respiratory activity for instance, but should come from H2O2. Only in strong saturating light (2 · 104 lx white light) an appreciable l8O2-uptake was observed. This uptake was sensitive to DCMU, o-phenanthroline, the uncoupler CCCP, KCN and SHAM. The uncoupler CCCP stimulated 18O2 uptake in strong light, but also affected as an ADRY-reagent photosynthetic oxygen evolution. CCCP abolished also the 1802-evolution signal in weak light (80 lx) demonstrating that the S-state system is involved in 18O2 production from H2O2. KCN also abolishes the weak light 18O2 evolution and inhibits 18O2 uptake in strong light. It appears that the low light intensity signal, also when measured as flash-induced signal, is inverted by KCN to an uptake phenomenon which, when analyzed under flash light, exhibits a periodicity of 2. It is suggested that part of the KCN effect observed with such preparations is due to an effect on cytochrome b-559. The effect of SHAM on oxygen uptake in strong light seems to indicate that O.chalybea owns in addition a CN−-insensitive respiratory pathway. Hence, in O. chalybea our method permits the distinction of at least three types of oxygen uptake: one is associated with Photosystem II and the S-state system, whereas the two other types apparently belong to the respiratory pathway.