THE determination of the nucleotide sequence of two 5S ribosomal RNAs by Brownlee, Sanger and Barrell represents a major advance in our knowledge of the structure of ribonucleic acids1. This RNA contains only the four normal bases A, U, C and G, and so it is an ideal system on which to test some of the physico-chemical techniques known to be sensitive to the conformation of nucleic acids in solution2–4. There is a wide range of plausible secondary structures which can be devised for 5S ribosomal RNA. The model conformation suggested by Brownlee et al. contains a relatively small percentage of base paired residues. It is possible, however, to design other models in which a much larger number of nucleoside residues is formed into double strand helices. One such model is illustrated in a two dimensional projection in Fig. 1. The remarkable property of this hypothetical conformation is the large fraction of residues in uninterrupted double strands. The number of residues involved in base pairing varies from 82 to 98 out of a total of 120. depending on the identity of the base in position 13 and on whether G–U pairs, two single base pairs and one loop with only two residues are included. The base composition of the 5S ribosomal RNAs would permit a theoretical maximum of 112 base paired residues. In our model as much as 87 per cent of this maximum is attained. This is a much larger percentage of double strand sections than seems possible for any of the transfer RNAs of known sequence5–8.