The origin and evolution of viruses is coming under stronger attack, thanks to the increasing availability of gene and genome sequences. However, the methodology is still narrow, being mainly based on alignment of sequences. In this paper I continue the effort to develop indexes for characterizing nucleic acid sequences and establishing relatedness among them [l]. As previously shown, the DNA sequences of papova virus SV40 and the untranslated zone following the chicken ovalbumin gene are similar in having very low frequencies for the dlnucleotide CG [I]. An additional likeness is here described in ‘poly(A) tendency’. A sequence with poly(A) tendency has an elevated frequency for purely adenine (A) containing o~gonucleotides of 4 or more bases in length. Two sample sequences high in runs of A are compared to all pub~~ed mRNA sequences [2] and to a large number of untranslated sequences. AAAA is the most frequent of the 256 possible tetranucleotides in the 637 untranslated bases 3’ to the chicken ovalbumin codons [3] (6370~3’) and in 4 papova virus genes, as revealed below. Although I find poly(A) tendency throughout the SV40 and BKV genomes [4-71, the VP1 gene is most like 6370~3’ in this respect. The maximum tendency in SV40 (the two papova genomes are very similar in poly(A) tendency, see below) is reached in the first half of VP1 .