S tudies utilizing in vitro microperfusion were designed to examine whether urea is actively or passively transported across superficial and juxtamedullary straight segments of rabbit proximal tubules. With perfusate and bath solutions containing 1 mM urea and electrolytes similar to normal plasma, the efflux (lumen-to-bath) isotopic permeability (X 10(-5) cm s-1) of superficial segments was 1.37 +/- 0.16 and of juxtamedullary segments was 2.14 +/- 0.20. In the same tubules, the influx (bath-to-lumen) isotopic permeability was 3.70 +/- 0.35 in superficial segments and 4.75 +/- 0.37 in juxtamedullary segments. Despite net water movement in the opposite direction (0.5 nl mm-1 min-1), the influx rate was significantly higher than the efflux rate of urea in both groups. With a low perfusion rate (2 nl/min) and equivalent specific activities of [14C]urea in bath and perfusate, the collected-to-perfused ratio of [14C]urea, corrected for volume marker change, was 1.07 +/- 0.01 in superficial and 1.09 +/- 0.01 in juxtamedullary nephrons, thus indicating net secretion in both segments. In separate studies urea influx was inhibited by hypothermia (decrease from 37 degrees to 28 degrees C), by phloretin (0.1 mM in bath), by cyanide (1 mM), but not by probenecid (0.2 mM). In each case the inhibition was highly significant and reversible. These data suggest that urea is actively secreted by the straight segments of both the superficial and juxtamedullary proximal tubules. These segments may, therefore, contribute significantly to the high urea concentration found at the bend of Henle's loop by micropuncture.