Seventy-three patients with chronic renal failure who underwent parathyroidectomy between March 1978 and April 1987 were reviewed. Thirty-four patients had undergone sub-total parathyroidectomy, and 39 patients had undergone total parathyroidectomy with parathyroid autograft into the forearm. Eight patients showed persistent hyperparathyroidism requiring a second surgical procedure. In all other patients there was highly statistical improvement in parathyroid hormone, total calcium, ionized calcium, alkaline phosphatase and a significant reduction in calcium x phosphate product. Histological evidence of osteitis fibrosa was present in 21 of 22 patients before surgery. Postoperatively, four showed complete resolution and improvement. Three patients developed histological evidence of osteomalacia during the study period. Only four of the 39 patients who underwent total parathyroidectomy with autograft had true recurrent hyperparathyroidism and only two of the 34 patients who underwent sub-total parathyroidectomy had recurrent disease, indicating that there is little to choose between the two techniques in the control of secondary hyperparathyroidism and its subsequent recurrence. In one patient with recurrence of hyperparathyroidism from a forearm parathyroid graft the histological picture was different from that of normal hyperplastic parathyroid tissue. Although it is probable that abnormal parathyroid tissue had been implanted there was no evidence of invasive growth into the forearm muscle. The most striking feature of long term follow-up was the difference in calcium x phosphate product in patients in whom vascular calcification increased compared to those patients with no change or regression of calcification. Mean calcium phosphate product in those patients with progressive vascular calcification was 4.93 for small and medium size vessels and 5.38 for large vessels compared to 4.10 for small and medium vessels and 4.09 for large vessels. In the former case the serum phosphate was 2.00 and 2.17 as compared to 1.75 or 1.73, suggesting that the aim in patients with end stage renal failure maintained by dialysis should be to control the serum phosphate concentration to 1.8 mmol or less and the calcium x phosphate product to less than 4.2.