To examine endogenous factors affecting the development of the massive bronchoconstriction in the postmortem guinea pig lung, 58 anesthetized open-chest animals were divided into three groups: 1) exsanguination only (n = 13), 2) pulmonary perfusion with 5% dextran and 1% bovine serum albumin (BSA) in Tyrode's solution (Ca2+ perfusate) (n = 21), and 3) pulmonary perfusion with 5% dextran and 1% BSA in saline (Ca2+-free perfusate) (n = 24). These groups were further divided into several subgroups according to treatments: 1) substance P depletion by chronic administration of capsaicin, 2) acute capsaicin treatment to release substance P, 3) dazoxiben treatment to block endogenous synthesis of thromboxane A2, 4) diethylcarbamazine treatment to eliminate leukotriene (LT) synthesis, and 5) FPL 55712 treatment to antagonize actions of LT. Vital capacity from the deflation pressure-volume (PV) curve of the lung was used as the indicator of bronchoconstriction. Most PV curves were performed for 30 min following exsanguination or artificial perfusion. Ca2+-free perfusate enhanced the airway spasm at 5-10 min, but the spasm disappeared gradually after 10 min. Substance P depletion significantly decreased (P less than 0.01) the bronchial constriction at 20-30 min, whereas substance P release induced severe airway spasm (P less than 0.01) during the entire study. In addition, FPL 55712 reduced the bronchospasm (P less than 0.05) in Ca2+ perfusate at 30 min. Thus Ca2+ and several endogenous mediators may be involved with the airway spasm of the postmortem guinea pig lung.