This study was designed to determine whether learning disabilities can be predicted from kindergarten teachers' anecdotal reports. Thirty-seven boys, presently diagnosed as learning disabled, were selected and their kinder-garten classes identified. Controls were chosen from the same kindergarten classes: boys of similar ages and IQs. Judges rated the reports on these boys with two types of measures. The first included seven categories of traits: immaturity, maturity, attention span, motor control, social and emotional adjustment, impulsivity, and speech and language. The second type included quantitative rating scales of general behavior common to the classroom. The two groups differed significantly on all scales of general behavior, and in all but one of the trait categories. The results are taken as evidence for the value of this approach of using informal kindergarten reports.