THERE is much interest in the development of 'designer ribozymes' to target destruction of RNAs in vitro and in vivo. Engineering of ribozymes with novel specificities requires detailed knowledge of the ribozyme-substrate interaction, and a rigorous evaluation of sequence specificity. The hairpin ribozyme catalyses an efficient and reversible site-specific cleavage reaction. We have used mutagenesis and in vitro selection strategies to show that RNA cleavage and ligation has an absolute requirement for guanosine immediately 3' to the cleavage-ligation site. This G is not required for efficient substrate binding, rather, its 2-amino group is an essential component of the active site required for catalysis.