Forty-eight handwriting specimens were rated by 6 handwriting analysts on 16 personality variables. Correlations of analysts ratings and personality inventory scores were nonsignificant. Analysis of variance showed that there were significant differences between the analysts on their judgments of 10 of the 15 traits. Subjects were also given the reports prepared for them and a bogus report and asked to rate the analysts' assessments for accuracy. Subjects were able to choose 2 of the 15 traits with significant accuracy on the real report when considering individual traits. There were no significant differences between the accuracy ratings on the real and bogus reports on overall rating. It was concluded that the analysts could not accurately predict personality from handwriting.