Studies comparing objective measures of sucking with data from finegrained clinical assessments of the neonate have shown significant correlations between painstaking and time-comsuming clinical methods which may only be reliably applied by highly trained clinician-investigators, and the data generated by a simple technique which can be rapidly and precisely administered in the nursery by nurses or technicians. Within a few minutes the sucking instrument can generate data that explain 50% or more of the variance in certain relevant factors of the Brazelton neonatal neurobehavioral assessment scale, which in our hands requires the participation of two trained clinician-investigators for a period of almost one hour for each test and recording session. There are certain limitations to the information directly available from the sucking measures. Clinical observations must be made in order to correctly interpret some of the findings such as the biphasic relationship between irritability and sucking. For example, an infant may not suck at all because it is obtunded, or it may not suck because it is overexcited. In the case of irritability, sucking performance provides a measure of the magnitude, but not of the polarity of the CNS arousal sucking correlates directly and gives a good estimate of both polarity as well as amount of these behaviors. Objective measures of sucking behavior are a convenient and reliable means for measuring drug effects in the nursery and may be useful in regulating therapy of the newborn.