The urine concentrating effect of the mammalian renal inner medulla has been attributed to countercurrent multiplication of a transepithelial osmotic difference arising from passive absorption of NaCl from thin ascending limbs of long loops of Henle. This study assesses, both mathematically and experimentally, whether the permeability criteria for effective function of this passive hypothesis are consistent with transport properties measured in long loops of Henle of chinchilla. Mathematical simulations incorporating loop of Henle transepithelial permeabilities idealized for the passive hypothesis generated a steep inner medullary osmotic gradient, confirming the fundamental feasibility of the passive hypothesis. However, when permeabilities measured in chinchilla were used, no inner medullary gradient was generated. A key parameter in the apparent failure of the passive hypothesis is the long-loop descending limb (LDL) urea permeability, which must be small to prevent significant transepithelial urea flux into inner medullary LDL. Consequently, experiments in isolated perfused thin LDL were conducted to determine whether the urea permeability may be lower under conditions more nearly resembling those in the inner medulla. LDL segments were dissected from 30-70% of the distance along the inner medullary axis of the chinchilla kidney. The factors tested were NaCl concentration (125-400 mM in perfusate and bath), urea concentration (5-500 mM in perfusate and bath), calcium concentration (2-8 mM in perfusate and bath), and protamine concentration (300 micrograms/ml in perfusate). None of these factors significantly altered the measured urea permeability, which exceeded 20 x 10(-5) cm/s for all conditions. Simulation results show that this moderately high urea permeability in LDL is an order of magnitude too high for effective operation of the passive countercurrent multiplier.