Pharmacologic study is needed in the extremely immature newborns who currently survive. Study is needed of both the drug treatment previously established in more mature neonates and of novel drug therapy. Carefully controlled studies are needed to identify accurately both beneficial and harmful drug therapy and the mechanisms of that toxicity. Careful pharmacologic study of drug disposition and its mechanisms might lead to dosing paradigms or patient selection that minimize toxicity and maximize efficacy. In vivo, translational models of neonatal diseases are limited, but can be used to identify novel treatments and study mechanisms of established, successful therapy. Findings from such studies can generate hypotheses for study in humans leading to a continuing scientific interchange from bedside to bench to bedside. Similarly, clinical observations can generate hypotheses for study in translational models where more invasive analyses are possible. Specific areas of drug treatment should focus on neonatal disorders with long-term, adverse outcomes, such as chronic lung disease, that is amenable to translational study with animal models. National data show a progressive decrease in the clinician-scientist pool entering biomedical research. The future of neonatal pharmacology studies requires an increase in training programs for the physician-scientist whose clinical education in neonatology can be complemented by rigorous basic-science training. Success as a clinician-scientist will require collaboration with full-time basic scientists who can continue studies during periods of clinical work and provide critical study methodology to the overall study design. Such a work environment must be supported by academic institutions and may require more flexibility in the promotion and tenure schedule and process, such as the nature of what it rewards. To complement this, the NIH could modify its grant reporting process to identify co-investigators in studies who may provide unique input to the study concepts and design, such as clinical correlations or clinical investigations.Journal of Perinatology (2006) 26, S8-S12. doi:10.1038/sj.jp.7211423. [ABSTRACT FROM AUTHOR]