Daily ultrasonographic measurements of transplant cross-sectional area were used to quantify allograft swelling as a diagnostic test for acute rejection in a series of 120 renal transplants. Initial graft function (IF) occurred in 86 patients (72%) and primary non-function (PNF) occurred in the remaining 34 (28%). An increase in allograft cross-sectional area greater than or equal to 10% was defined as a positive ultrasound scan suggesting an acute rejection episode and was investigated by needle core biopsy. During periods of PNF, allografts with consistently negative ultrasound scans were submitted to needle core biopsy on a weekly basis. The diagnosis of rejection was based exclusively on the histological findings. In the IF group, agreement between ultrasound and histological diagnosis was good (k = 0.63, sensitivity 81%, specificity 83%, positive predictive value 76%, negative predictive value 86% and overall accuracy 82%). In the PNF group, agreement between ultrasound and histology was only fair (k = 0.46, sensitivity 77%, specificity 70%, positive predictive value 69%, negative predictive value 78% and overall accuracy 73%). It is concluded that a degree of allograft swelling is sometimes associated with acute tubular necrosis, and this makes ultrasound measurements of transplant size a less useful technique of monitoring kidneys with PNF.