The bromocriptine growth hormone test (BGHT) was used to monitor D2 receptor activity in a group of 16 chronic schizophrenics who during the baseline phase were receiving greater than or equal to 20 mg/day haloperidol. In all subjects at baseline, the rise in plasma GH in response to the oral administration of bromocriptine (50 micrograms/kg) was blocked. The dose of haloperidol was then gradually reduced; the BGHT was repeated as each new dose was established. No escape from blockade of the GH response was observed until the dose of haloperidol was lowered to 10 mg/day (3 of 16 subjects escaped from blockade). At this dose the average plasma haloperidol level was 4 ng/ml. Two additional subjects escaped as the dose was reduced to 5 mg/day and six more escaped as the dose was reduced to 2.5 mg/day. The average haloperidol plasma level at 5 and 2.5 mg/day was 1.6 and 1.2 ng/ml respectively. The remaining five subjects escaped from blockade as the dose was reduced to 0 mg/day. In five subjects, escape from blockade was associated with a significant decrease in positive psychotic symptoms; in these subjects reestablishing the "just" blockade dose of haloperidol did not increase psychotic symptoms. In nine subjects escape from blockade was associated with an increase of positive psychotic symptoms; in six of these patients, reestablishing the "just" blockade dose of haloperidol attenuated psychotic symptoms to near baseline levels. We conclude that the GH challenge test is a useful adjunct to dose-reduction in the chronic patient. Furthermore, for some patients the "just" blockade dose appears to be near the minimum dose with the maximum therapeutic effect.