A study undertaken in 1996 of Australia"s performance in the high impact journals of a few selected fields of science has produced empirical data for examining the factors that influence peers in their choice of the "highly-rated" journals in their field. A number of characteristics were used to compare the selected journals with those having the highest impact factor, as listed in ISI"s Journal Citation Reports. This paper ranked journals on three impact factors – ISI"s impact factor for two consecutive years, and one calculated for a five-year window. The data suggests that the type of impact measure was less important in journal selection than the long-term validity of the rankings. A group of experts was less likely to include journals that were only highly ranked for a short period in their "top 20". Of the more descriptive journal characteristics analysed, the age of the journal appeared significant. Their selections also appeared biased against journals that were relatively new, regardless of how high their impact factor was.