Medullary thick ascending limbs of Henle's loop of the Swiss-Webster mouse were perfused in vitro with an isotonic perfusate and a Ringer's bathing medium. In five studies, addition of a supramaximal concentration of synthetic arginine vasopressin (AVP) to the bathing medium resulted in an increase in electrical potential difference (PD) from 5.0 +/- 1.5 mV, lumen positive, to 10.7 +/- 1.4 mV (P < 0.001). When AVP was removed, the PD returned to 2.6 +/- 0.9 mV (P < 0.001), then increased again to 6.9 +/- 1.7 mV (P < 0.01) when AVP was added a second time. A significant, but submaximal, increase in PD of 2.3 +/- 0.6 MV (P < 0.05) was observed in five medullary thick ascending limbs when AVP was added to the bathing medium at a concentration of 10 microunits/ml. This increase was approximately one-third of the response observed at a concentration of 100 microunits/ml in the same tubule. No further increment in PD was observed in five medullary thick ascending limbs when the AVP concentration was increased from 100 to 1,000 microunits/ml. In seven thick ascendcing limbs, the effect of AVP on PD was reproduced by the addition of 8-[p-chlorophenylthio]-cyclic 3',5'-adenosine monophosphate to the bathing medium at a final concentration of 0.1 mM. AVP increased unidirectional chloride flux from lumen to bath from 29.3 +/- 3.2 to 69.8 +/- 6.2 peq/cm per s (P < 0.001) in spite of an increase in the lumen positive PD from 1.6 +/- 0.5 mV to 7.0 +/- 0.6 mV (P < 0.001). Unidirectional chloride flux from bath to lumen was not affected by AVP. In another series of experiments, net chloride flux increased from 15.6 +/- 3.0 to 41.7 +/- 5.3 peq/cm per s (P < 0.05) after addition of AVP. The effect of AVP on hydraulic water permeability (Lp) was examined by adding raffinose to the bathing medium in both the presence and the absence of AVP. The calculated Lp of 16 +/- 2 nm/s per atm in the absence of AVP, although very low, was significantly different from zero (P < 0.01). However, the Lp did not increase significantly when AVP was added to the bathing medium. These results suggest that AVP has a second site of action in the kidney to increase chloride transport by the medullary thick ascending limb in addition to its well-known effect on the water permeability of the collecting tubule. The former effect would contribute to urinary concentrating ability by increasing the axial osmotic gradient in the renal medulla.