Eighty-one preterm infants (mean gestational age 29 weeks, range 24-36 weeks) discharged from The Johns Hopkins Hospital Neonatal Intensive Care Unit were followed up sequentially from birth to 2 years of age by use of the Clinical Linguistic and Auditory Milestone Scale (CLAMS) to evaluate language development. Children were studied during three time intervals: Interval 1: 3-5 months chronologic age (CA); Interval 2: 9-14 months (CA); and Interval 3: 18-24 months (CA). Psychometric test scores were compared with CLAMS Language Quotients (LQ) by use of full, partial (75%, 50%, 25%), and no "correction" for weeks of prematurity to determine whether "correcting" for prematurity would yield a more accurate estimate of eventual cognitive outcome. CLAMS LQ at Interval 1 was highly correlated with CLAMS LQ at Interval 2 and CLAMS LQ at Interval 2 correlated well with CLAMS LQ at Interval 3 (r = 0.57 and 0.64, respectively, P = 0.0001). Correlations indicated that there was an orderly, sequential development of language in the preterm infant. CLAMS evaluations correlated significantly with psychometric test results during Interval 2 and Interval 3 (r = 0.34, P < 0.02 and r = 0.75, P = 0.0001, respectively). The CLAMS proved to be a useful instrument for monitoring preterm language development in the primary pediatric care setting.