A radioimmunoassay for serotonin (5-hydroxytryptamine) has been developed, validated, and applied to the measurement of serotonin in blood and platelet-rich plasma. Six rabbits immunized with serotonin diazotized to a DL-p-aminophenylalanine-bovine serum albumin conjugate yielded anti-serotonin antibodies. In the radioimmunoassay, antibody-containing plasma (1:100) is incubated with 0.2 pmoles of [3H]serotonin, EDTA, and either serotonin standards or unknown samples (0.1 ml). Blood levels of serotonin are measured in a protein-free supernatant prepared by water lysis of heparinized blood followed by protein precipitation using zinc hydroxide. This assay is sensitive to 100 pg of serotonin and has demonstrated insignificant cross-reactivity with a number of serotonin analogues at their normal circulating concentrations. Validation has been achieved by obtaining comparable values for normal blood serotonin concentrations by radioimmunoassay and by spectrophotofluorometry as well as by demonstrating that dilutions of endogenous serotonin in rabbit blood and blood from a patient with the carcinoid syndrome were superimposable on a standard calibration curve. In 55 normal human subjects the mean whole blood serotonin concentration was 168 +/- 13.4 ng/ml (mean +/- SEM) (range: 31 to 442 ng/ml). In 15 normal volunteers the mean radioimmunoassayable serotonin concentrations in whole blood and platelet-rich plasma were 337 +/- 40 ng/10(9) platelets and 341 +/- 37 ng/10(9) platelets, respectively. Incubation of blood with PGE1 to inhibit in vitro platelet aggregation before radioimmunoassay resulted in a significant fall in measurable serotonin activity in platelet-poor plasma (from 15.3 +/- 3.0 to 6.4 +/- 1.2 ng/ml). Seventeen normal human volunteers demonstrated a rise in circulating serotonin activity to a mean of 362.1 +/- 16.9 ng/ml at 30 min postcibal after a standard test meal, which was significantly (P less than 0.02) greater than the mean fasting level of 198.1 +/- 37.0 ng/ml. Five fasting controls did not show a rise in circulating serotonin levels when sampled at these intervals. These data suggest release of serotonin, presumably from the intestine, after a meal and make serotonin a candidate hormone in gastrointestinal physiology.