Metabolic clearance rates (MCR) and production rates (PR) of prolactin (PRL) have been determined by the constant infusion to equilibrium technique in 11 normal subjects, 6 patients with hyperthyroidism, 4 patients with hypothyroidism, and 9 patients with hyperprolactinemia. PRL MCR was also determined tin four patients during dopamine infusion. Mean PRL MCR was 46 +/- 1 ml/min per m2 in women and 44 +/- 3 ml/min per m2 in men, and was significantly correlated with body mass (r = 0.84, P less than 0.001). In contrast with controls, PRL MCR was higher in hyperthyroidism (MCR = 52 +/- 8 ml/min per m2, P less than 0.05), was slightly lower in hypothyroidism (MCR = 38 +/- 10 ml/min per m2, P = NS), and was significantly correlated with serum thyroxine (r = 0.46, P less than 0.02). PRL MCR was lower than controls in hyperprolactinemia (MCR = 40 +/- 5 ml/min per m2, P less than 0.01) and was inversely correlated with serum PRL (r = -0.72, P less than 0.001). PRL MCR was not significantly changed by dopamine infusion. Mean PRL PR for women and men was 211 +/- 74 and 187 +/- 44 micrograms/d per m2, respectively (P = NS). In hyperthyroidism the PRL PR was elevated (PR = 335 +/- 68 micrograms/d per m2, P less than 0.02), but in hypothyroidism the increase (PR = 233 +/- 159 micrograms/d per m2) was not significant. In hyperprolactinemia the PRL PR was extremely high (PR = 31,000 +/- 29,000 micrograms/d per m2). Dopamine infusion decreased RPL PR from 270 to 66 micrograms/d per m2 indicating that its effect was on pituitary PRL secretion and not PRL metabolism. To evaluate possible circulating PRL heterogeneity that might arise during infusion, gel filtration of infusate and serum obtained during the MCR procedure was performed. Labeled monomeric PRL (peak III, Kav (partition coefficient) = 0.4) was partially converted to two larger forms (peaks I and II) in vivo. Peak I (Kav = 0) was 30--40% immunoprecipitable, although peak II (Kav = 0.2) was not immunoprecipitable. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of peak I resulted in greater than or equal to 90% conversion to peak III and restoration of full immunoactivity. Thus, peak I is a noncovalently linked aggregate that is partially immunoactive, and therefore able to alter MCR determinations. These studies demonstrate the impact of hormone heterogeneity on MCR estimations and suggest that gel filtration of immunoprecipitable material be an integral part of future MCR measurements.