The value of thrice weekly technetium-99m mercaptoacetyltriglycine renography after renal transplantation was investigated in 213 consecutive transplants. A grading system was used: 0 = normal renogram; 1 = normal uptake, reduced excretion; 2 = normal uptake, flat excretion curve; 3 = rising curve; 4 = reduced rate of uptake, rising curve and reduced absolute uptake; 5 = minimal uptake. The initial renogram grade (RG) was primarily a marker of ischaemic damage, being poorer with cadaver donation, long cold ischaemia (>24 h), and high donor and recipient age. High primary RG predicted primary graft non-function, long time to graft function, low discharge Cr EDTA clearance and low 1- and 5-year graft survival. Discharge RG predicted late (>6 months) graft loss. RG was highly correlated (P<0.001) with creatinine and creatinine clearance, and changes in RG were correlated with changes in renal function. A change in RG of 0.5 was non-specific, while a change of 1 or more predicted clinical complications in 95% of cases. The negative predictive value was low (58%). RG change antedated clinical diagnosis in only 38% of cases, and in only 14% of acute rejections did an RG change of 1 or more antedate a rising creatinine. RG did not contribute to the differential diagnosis between acute rejection, acute tubulointerstitial nephropathy and cyclosporine toxicity. In conclusion, an initial renography after transplantation is valuable as it measures ischaemic damage and predicts duration of graft nonfunction and both short and long-term graft survival. A review of the literature suggests that the indication for serial scintigraphic monitoring for functioning grafts is less certain: the diagnostic specificity is insufficient for it to be the definitive investigation for common diagnostic problems and it does not give sufficient advance warning of impending problems.