The measurement of neonatal responses to painful stimuli remains a significant clinical problem. Although numerous measures have been evaluated, instruments that are valid, reliable, and clinically feasible are not yet available. The purpose of this paper is to critique the studies that have been done using biochemical, physiological, and behavioral measures to evaluate neonatal responses to painful stimuli. Specific issues regarding measurement in premature and critically ill neonates are emphasized. The intent of this review and critique of the literature is to stimulate additional research into the assessment of neonatal pain.