Clinical complications of transfusional iron overload are still common in patients with thalassaemia major (TM) and it is not clear how best to monitor body iron stores during long-term follow-up to anticipate tissue damage. In this study, we have reviewed a group of 32 patients who underwent liver biopsy between 1984 and 1986. We developed a method of assessing the trend in serum ferritin (TSF) during long-term monitoring and compared this with mean serum ferritin (MSF) and initial liver iron (LI) concentration to determine whether, individually or in combination, they were accurate in predicting clinical outcome. LI levels were low (< 7 mg/g), medium (7-15 mg/g) and high (> 15 mg/g dry weight) in 15, 7 and 10 patients respectively. MSF was low (< 1500 microg/l), medium (1500-2500 microg/l) and high (> 2500 microg/l) in 10, 14 and 8 patients. TSF was low, medium and high risk in 9, 9 and 11 out of 29 evaluable patients. During a median follow-up of 13.6 years (range 2.3-14.8 years) after biopsy, nine patients died and an additional three patients developed heart failure. Hypothyroidism developed in five, hypoparathyroidism in four, and diabetes mellitus in seven patients. Cirrhosis developed in four of 10 evaluable patients. The clinical end-point of death or cardiac failure was significantly associated with increasing iron load using all three means of assessment. Although numbers were insufficient for statistical analysis, MSF or TSF were more closely associated with complications of iron overload than LI. There was no clear additional value in combining LI with MSF or TSF. The data show that quantitation of liver iron from a single liver biopsy has little value in long-term monitoring of iron stores. Most complications can be avoided if ferritin levels can be brought down to <1500 microg/l.