The effects on porcine arterial structure and permeability of a 4-hour in vitro incubation at 37°C in eight different blood-derived and synthetic nutrient media were examined. Changes in arterial permeability were inferred from the normalized, 1-hour, pressurized, transendothelial uptake (M/c t cm) of porcine l25 I-albumin in 60 porcine aortic tissue preparations using an organ-support system. The organ-support system provided experimental control of ambient gas composition, temperature, transmural pressure, flow (stirring), and nutrient media at a number of sites along the vessel. Light and electron microscopic (scanning and transmission) structural correlations with the observed permeability changes were examined. The M/c 0 from the autogenous serum (AS) medium was used as the "control" measurement in each vessel preparation. (Grand mean M/c 0 for AS from all studies was 0J12±0.011 [xlO~3] [mean±SEM] cm, n=60.) For brevity, M/c 0 values from the other media are expressed below as a percentage of the corresponding paired Af/c, from the AS. Uptake from heparinized autogenous blood was 113±9% of that from AS (p=0.119); from heparinized autogenous plasma was 135± 10% (p=0.048); from AS+heparin was 97±5% (p=0.498); from pooled porcine serum was 113±9% (p=0.037); from a synthetic medium was 131 ±8% (p=0.004); and from a physiological hetastarch solution was 532±8% (p=0.0002). Associated light microscopic structural changes and ultrastructural changes were not found. We conclude that 1) incubation with AS and heparinized blood (both of which are autogenous blood substances containing platelet products or platelets) provided the best support for the endothelial barrier function, whereas heparin plasma, pooled serum, a synthetic medium, and particularly hetastarch provided poorer support; 2) arterial permeability can increase significantly without discernible endothelial ultrastructural changes; and 3) AS and to a lesser extent heparin blood should provide a suitable nutrient medium for short-term (<4-hour) metabolic support of the endothelial surface and subjacent tissues. A understanding of the essential factors required for optimum metabolic support of excised blood vessels in vitro is of considerable interest in many areas of vascular biology. In studies of basic vascular biology, the development of an appropriate, metabolically supported, in vitro arterial preparation would provide an important experimental bridge between tissue-culture methodology and in vivo experimentation. For example, unlike those in a tissue-culture system, the cells in such an organ-support system (OSS) would continue to communicate with their normal matrix milieu and cellular cohorts. Unlike the in vivo state,