In order to probe the interaction between tRNA and the tRNA hypermodifying enzyme, tRNA-guanine transglycosylase (TGT) from Escherichia coli, we have undertaken the generation of E coli tRNA(Tyr) and analogues. During efforts to adapt currently available in vitro, transcription techniques we encountered difficulties attributable to dimerization of the tRNA products. E coli tRNA(Tyr) has previously been characterized for its ability to form a dimer in solutions of suitable salt concentrations at appropriate temperatures (Yang SK, Soll DG, Crothers DM (1972) Biochemistry 11, 2311-2320; Rordorff BF, Kearns DR (1976) Biochemistry 15, 3320-3330). We have applied similar techniques to our unmodified analogue of E coli tRNA(Tyr) and produced both monomeric and dimeric forms of E coli tRNA(Tyr). In this report we find that the dimer does serve as a substrate for modification by TGT. While both the conformers are equal in terms of V-max (within experimental error) a 2.5-fold increase in K-M occurs when going from monomer to dimer. This suggests that TGT preferentially binds the monomer but once either conformer is bound will catalyze the modification reaction equally well. We have also compared the results for the two conformers to our previous data of an RNA minihelix corresponding to the anticodon arm of E coli tRNA(Tyr). Here we find that our earlier conclusion, that the recognition elements for TGT are localized within the anticodon arm of cognate tRNAs, is supported.