Researchers have determined that the lay public makes risk judgments in ways that are very different from those advocated by scientists. Noting that these differences have caused considerable concern among those who promote and regulate health and safety, a study examined media coverage of the accident at the Robert E. Ginna nuclear power plant near Rochester, New York in 1982. It was hypothesized that (1) publications whose targeted audiences are more scientifically sophisticated will publish more specific risk assessment information, and (2) publications whose targeted audiences are less scientifically sophisticated will publish significantly more general risk assessment information. Publications in five categories, including newsmagazines for scientists, popular science magazines, prestige newspapers, news magazines, and newspapers were analyzed for such key variables as the words "risk,""risk estimation,""general risk statement," and "specific risk statement." As expected, results showed that publications whose audiences are more scientifically sophisticated printed more specific risk assessments than did those aimed at general audiences. However, articles in scientists' magazines used comparisons with greater frequency than did those in publications aimed at general audiences, and scientific publications were no more likely than nonscientific publications to cite sources for risk assessment information. Reference notes and six tables are appended. (DF)