In 66 untreated patients with hyperthyroidism, serum triiodothyronine (T(3)) and thyroxine (T(4)) concentrations were measured by immunoassay. The mean T(3) level was 478+/-28 ng/100 ml (all values mean+/-SEM) and the T(4) was 20.6+/-0.6 mug/100 ml. The serum T(4)/T(3) ratio by weight was 48+/-2 as opposed to a value of 71+/-3 in euthyroid adults. There was a significant inverse correlation of the T(4)/T(3) ratios with serum T(3) (r=0.77; P<0.01) but not with serum T(4)(r=0.21). These results suggested that relative overproduction of T(3) is consistently present in patients with hyperthyroidism. To examine the acute effects of various antithyroid agents on serum T(3) and T(4) concentrations, iodide, propylthiouracil (PTU), and methylmercaptoimidazole (MMI) were given alone to mine patients, and serial T(3) and T(4) measurements were made. There was an acute decrease in serum T(3) over the first 5 days in the three iodide and three PTU-treated patients which was greater than that seen in the MMI group. This suggested that PTU and MMI had different effects on T(3) production. To compare the effects of PTU and MMI under conditions in which thyroidal hormone release was minimized, these drugs were given in combination with iodide. The mean daily dosage of PTU was 827 (n=11) and of MMI was 88 (n=8). In the PTU+iodide group, the initial serum T(3) concentration was 586+/-61 ng/100 ml and decreased significantly to 326+/-41 on day 1 and to 248+/-21 on days 2 and 3, respectively, and did not change further on days 4 and 5. In the MMI + iodide group, basal serum T(3) was 645+/-90 ng/100 ml and decreased to 568+/-81, 452+/-73, and 344+/-51 on days 1, 2, and 3, respectively, and did not change thereafter. While the initial T(3) concentrations in serum were not different in the PTU and MMI groups, the T(3) concentrations in the PTU patients were significantly lower on days 1 and 2 and during the apparent plateau period on days 3-5. Serum T(4) concentrations decreased gradually in both groups, from 23.9+/-2.0 mug/100 ml, initially, to 17.5+/-1.6 on day 5 in the PTU group and from 22.0+/-2.6 to 14.6+/-2.0 in the MMI-treated patients. The T(4) values were not significantly different at any time. These changes resulted in increases in the serum T(4)/T(3) ratios in both groups, but these ratios were substantially higher in the patients treated with PTU + iodide. The initial serum T(4)/T(3) ratio was 43+/-3 and increased to 74+/-7 and 88+/-7 on days 1 and 2 in the PTU group, reaching a plateau value of 91+/-7 during days 3-5. Comparable values for MMI-treated patients were 35+/-2, 42+/-3, 52+/-6, and 54+/-3 during the plateau period. Previous investigations have shown that PTU inhibits T(4) deiodination in hyperthyroid patients and decreases T(3) production from T(4) in animals. The greater acute decrease in serum T(3) and the higher serum T(4)/T(3) ratios in the PTU-treated patients seems best explained by an inhibition of peripheral T(3) production by this agent. This conclusion is further supported by a direct relationship between the T(4)/T(3) ratio on days 3-5 and the dose of PTU administered. These results further suggest that both thyroidal and extrathyroidal pathways contribute substantially to the apparent overproduction of T(3) in hyperthyroidism.