The long-term effects of two schedules of radioiodine therapy in patients with toxic multinodular goitre were evaluated. Forty-five patients (group A) were treated with low doses and 58 patients (group B) with calculated doses adjusted for thyroid weight (1.85-3.70 MBq/g) and radioactive iodine uptake. Follow-up (mean _+ SEM) was 4.3 + 0.2 years and 5.2 + 0.3 years, respectively (P>0.1). At the end of follow-up, hyperthyroidism was successfully reversed in 73% (group A) and 88% (group B). In each group, hypothyroidism was present in 7%. The total dose per gram of thyroid tissue was not significantly different in groups A and B (2.1 + 0.2 vs 2.7 _+ 0.2 MBq/g). However, for patients treated with calculated doses the number of 131I administrations was significantly lower (1.3 _+ 0.1) than for patients treated with low doses (2.2 _+ 0.2), and the percentage of patients who were adequately treated with a single dose was more than twice as high (66% in group B versus 27% in group A). Euthyroidism was reached within a significantly shorter time after treatment with calculated doses (median time 0.6 years in group B vs 1.5 years in group A; life table analysis). It is concluded that radioiodine is an effective treatment for toxic multinodular goitre with a low risk of post-treatment hypothyroidism and that calculated (higher) doses appear to be preferable to low doses.