Inositol was detected by gas-liquid chromatography in fluid perfused at 1.3 microliter/min through the lumen of the distal cauda epididymidis of anaesthetized rats. The concentration (247 microM) exceeded that in blood (48 microM) and the secretion rate was constant for 5 h. D-[3H]Myo-inositol infused for 3 h into the general circulation of rats (1 muCi/min) also appeared in fluid perfusing the lumen, whether or not 50 mM-inositol was present in the perfusing solution. No plateau of radioactivity was reached during infusion, and by the end of 3 h perfusate activity was 26% of that in blood. Calculation of the specific activity of inositol in perfusates and blood plasma suggested that blood was not the immediate source of luminal inositol, and that any endogenous pool of cyclitol has a turnover time of greater than 3 h. [3H]Myo-inositol perfused through the lumen was not absorbed by the tissue. These data suggest that the high concentration of inositol in epididymal fluid (49 mM) is derived in part by epididymal secretion from a pool that is only slowly replenished from blood, and maintained by the impermeability of the epithelium. Glucose also appeared in fluid perfused through the epididymal lumen, but its concentration (461 microM) was much less than in blood (8.5 mM), so this sugar may diffuse down a concentration gradient.