This study investigated the potential role of differential processing of internals and externals when making an attribution of responsibility about another for an automobile accident. Of the total subjects (96) half were defined as internals and the other half as externals. All subjects read an account of the accident with half the subjects reading a version that had relevant pause units (environmental and personal) underlined with spaces provided for these subjects to make importance ratings about the underlined units. All subjects filled out a questionnaire designed to assess attribution of responsibility and were required to recall the account. Results showed that the attempt to replicate previous research (Sosis, 1974) was not successful. This outcome severely limited any conclusions that could be made about the results from the underlining condition, which indicated no differences between internals and externals in terms of mean weightings for personal and environmental pause units. Also there were no observed differences between internals and externals for recall of personal and environmental units recalled to total units recalled, with the exception that externals recalled significantly more units than internals when recall followed the questionnaire. This difference was primarily due to the number of nonrelevant units recalled. Externals recalled significantly more nonrelevant units than internals when recall followed the questionnaire. Further attempts (Experiments I and II) to replicate the effect reported by Sosis were unsuccessful. Three possible explanations for not replicating the effect were advanced which included problems with (1) methodology, (2) the account of the accident, and (3) the sample of externals used. / Master of Science