Bacteria of the genus Methylobacterium are well-studied examples of facultative methylotrophs. These bacteria are classified as α-proteobacteria and are capable of growth on methanol and methylamine as well as on a variety of C2, C3, and C4 compounds (8). Methylobacterium strains are commonly found in soils, as well as on the surfaces of leaves of a wide variety of plants (1). Because of their distinctive pink pigmentation, they are sometimes referred to as PPFMs (pink-pigmented facultative methylotrophs). Evidence exists that these bacteria utilize methanol emitted by the stomata of plants (9), but the details of their relationship to plants has been unclear. Are Methylobacterium strains commensal bacteria, or do they communicate with plants in a more intimate relationship? A number of reports suggest that Methylobacterium strains are more than passive passengers on plant leaves. For instance, they have been shown in some cases to stimulate seed germination and plant development, possibly by production of phytohormones (2, 3, 4), and it has been reported that one strain produces the cytokinin zeatin (5) while others have been reported to produce indole acetic acid (6). In addition, Methylobacterium strains have been suggested to contribute to the flavor of strawberries (12) and have been localized as endosymbionts within cells of the buds of Scotch pine (Pinus sylvestris) (10). One nonpigmented Methylobacterium strain has been shown to form a root-nodulating nitrogen-fixing symbiosis with a legume (11). A partial genome sequence of Methylobacterium extorquens AM1 is available (http://pedant.mips.biochem.mpg.de/), which reveals a number of open reading frames with significant identity to genes involved in plant association in rhizobia and Agrobacterium (M. Lidstrom, unpublished results). Therefore, at this point numerous indications exist that these PPFMs interact with plants, but the biochemical and genetic details of those interactions have remained elusive. In this issue, Koenig et al. (7) present a set of studies that represents an important step in understanding the molecular basis of Methylobacterium-plant interactions. They show that four different Methylobacterium strains representing leaf isolates and a type M. extorquens strain all produce the cytokinin trans-zeatin at low levels in pure culture and excrete it into the culture medium. Biochemical and genetic evidence is presented suggesting that the trans-zeatin is not synthesized de novo but instead is derived from tRNA. This result is surprising, since so far the tRNA-derived zeatin from bacteria and plants has been restricted to the cis isomer. Is the low-level production of trans-zeatin observed in these strains sufficient to have an effect on plants? Surprisingly, the Methylobacterium-specific stimulation of seed germination that has been described in the past (3, 4) was unaffected in a mutant that is incapable of producing trans-zeatin. Thus, the answer to this question is still outstanding, as is the source of the stimulation effect itself. However, the results presented by Koenig et al. provide mechanistic evidence for cytokinin production by Methylobacterium strains, leading the way for future studies of the Methylobacterium-plant relationship. In addition, this evidence that a commensal bacterium produces a phytohormone via a previously unknown route provides new insights into the role of commensal plant bacteria in general.