In summary, little progress has been made in the past several years with respect to the treatment of the baby with BPD. The conduct of convincing clinical research seems to be a casualty of budget cuts and a rush to learn the tools of molecular biology. To date, there are no clinical trials that have convincingly demonstrated that long-term diuretic, bronchodilator, vasodilator, or antioxidant therapy is effective in the treatment of chronic BPD. Short-term corticosteroid therapy hastens extubation, but long-term outcome is unaffected and serious questions remain about its safety. Multicenter clinical trials should be carefully designed and implemented to address the values of these therapies. In the design of these trials, care should be taken to stratify treatment groups for known risk factors for BPD. What are the future directions for the treatment of BPD? It is hoped that new BPD treatment strategies will be based on an improved understanding of mechanisms of lung repair and inflammation. Enzyme, gene, cytokine, antioxidant, and antiprotease therapies are being developed in animal models of lung injury. In addition, the use of lung transplantation has begun to be explored for severe cases of BPD. It is also possible, as has occurred in many chronic idiopathic diseases, that nonspecific treatment may prove beneficial. Perhaps it is only a matter of time before intravenous immunoglobulin, cyclosporine, methotrexate, or "biological response modifiers" will be administered to infants with severe BPD. For example, there is anecdotal evidence that recombinant human growth hormone may improve respiratory muscle function in adults with chronic obstructive pulmonary diseases. In the absence of convincing clinical trials, the clinician should reserve existing therapies for the ventilator-dependent infant or infants whose high oxygen requirement is prohibiting discharge or resulting in complex home care or frequent rehospitalizations. It should be emphasized that continuous oxygen therapy combined with avoidance of environmental inhalant and infectious hazards have the strongest rationale and widest margin of safety for treatment of the infant with BPD. Ironically, oxygen therapy is frequently underutilized and discontinued too rapidly. Early discontinuation of oxygen therapy with alveolar hypoxia results in feeding difficulty, slow growth, nutrient malabsorption, bronchoconstriction, and pulmonary hypertension. Oxygen therapy, although more cumbersome and certainly less glamorous than other pharmacologic agents, remains the essential element of BPD care.