The reaction of hemoglobin with tetramethylbenzidine was studied with a centrifugal analyzer. We found that the hemochromogen reaction was significantly influenced by albumin. Increasing amounts of albumin apparently stimulate the hemochromogen reaction during the earlier part of an incubation and inhibit it later in the incubation. However, during an intermediate period, when the crossover from stimulation to inhibition occurs, the assay is independent of albumin and bilirubin concentrations. Increasing the temperature of incubation or the hydrogen peroxide concentration, or decreasing the tetramethylbenzidine concentration, shortens the crossover time. By adjusting these variables the method can be used with most instrumentation. We developed procedures for use on a centrifugal analyzer or for manual assay (reaction times, 4 and 20 min, respectively). Accuracy, as indicated by a comparison with a direct hemoglobin method (r = 0.990) and by recovery experiments, was excellent. The CV for the automated assay was less than 3%. Plasma collected in citrate and heparin showed good recovery of hemoglobin; recovery in EDTA was poorer.