This study investigated the diagnostic effectiveness of the Language Assessment Tasks (LAT) in differentiating language disordered students from nonhandicapped students. LAT was administered individually to two groups of subjects (experimental and control) matched for sex, race, and birthdate (ages 9-14 years). The performances of each group of subjects were analyzed by subtest items and for each language modality--comprehension, production, language content, and communicative competence. For data analysis and for grouping the subjects by language modalities, a numerical scoring system was developed. In all cases, the nonhandicapped subjects had higher performance scores. Not all subtests were useful in differentiating learning disabled subjects from the nonhandicapped, and many of the LAT items were equally difficult for learning disabled and nonhandicapped subjects. Correlation did not show a relationship between age and performance and sex and performance. Thirty-three percent of the test items were found to be effective in discriminating the language disabled students from the nonhandicapped students, and 17 percent were equally difficult for both groups of subjects. It was concluded that LAT can be utilized as a criterion referenced instrument with those items shown to differentiate the language disabled students from the nonhandicapped students. Modifications of scoring procedures are suggested. The specific findings are illustrated in four figures following the study. (JAZ)