Chenopods synthesize betaine by a two-step oxidation of choline: choline → betaine aldehyde → betaine. The pathway is chloroplastic; the first step has been shown in isolated spinach (Spinacia oleracea L.) chloroplasts to be O2- and light-dependent, the role of light being to provide reducing power (P Weigel, EA Weretilnyk, AD Hanson 1988 Plant Physiol 86: 54-60). Here, we report use of in vivo18O- and 2H-labeling in conjunction with fast atom bombardment mass spectrometry to test for two hypothetical choline-oxidizing reactions that would explain the observed requirements for O2 and reductant: a desaturase or an oxygenase. Simple syntheses for 2H3-choline, 2H3, 18O-choline, and 2H3, 18O-betaine are given. A desaturase mechanism was sought by giving choline deuterated at the 2-carbon, or choline unlabeled at this position together with 2H2O and by analyzing newly synthesized betaine. About 15% of the 2H at C-2 was lost during oxidation of choline to betaine, and about 10% of the betaine made in the presence of 50% 2H2O was monodeuterated. These small effects are more consistent with chemical exchange than with a desaturase, because 10 to 15% losses of 2H from the C-2 position also occurred if choline was converted to betaine by a purified bacterial choline oxidase. To test for an oxygenase, the incorporation of 18O from 18O2 into newly synthesized betaine was compared with that from 18O-labeled choline, in light and darkness. Incorporation of 18O from 18O-choline was readily detectable and varied from about 15 to 50% of the theoretical maximum value; the 18O losses were attributable to exchange of the intermediate betaine aldehyde with water. In darkness, incorporation of 18O from 18O2 approached that from 18O-choline, but in the light was severalfold lower, presumably due to isotopic dilution by photosynthetic 16O2. These data indicate that the chloroplast choline-oxidizing enzyme is an oxygenase.