In patients with chronic heart failure increased sympathetic activity is related to unfavourable prognosis. Since myocardial iodine-123 metaiodobenzylguanidine ([123I]MIBG) uptake is related to myocardial noradrenaline content, i.e. cardiac sympathetic activity, measurement of myocardial [123I]MIBG uptake may be of clinical use in determining prognosis or the effect of pharmacological intervention in these patients. The aim of the present study was to evaluate a new method to quantitate myocardial [123I]MIBG uptake with respect to reproducibility and accuracy. Eighteen [123I]MIBG planar and single-photon emission tomographic (SPET) studies of patients with chronic heart failure were evaluated. Myocardial uptake was calculated from the myocardial (MYO) to left ventricular cavity (C) count density ratio and the 123I activity in a blood sample. This was performed employing planar LAO images, a single-slice SPET method using the midventricular myocardial short-axis slice, and finally a multi-slice SPET method analysing semi-automatically drawn volumes of interest (VOIs). The accuracy of the multi-slice SPET method was verified using a cardiac phantom. The planar method was found to be reproducible [intra- and interobserver coefficients of variation (IACV and IRCV) were 0.025 and 0.012 respectively] but the mean MYO/C count density ratio was only 1.31 +/- 0.16 as a consequence of overprojection. For the single-slice SPET method IACV was 0.2 and IRCV was 0.13, representing poor reproducibility. For the multi-slice SPET method IACV was 0.051, IRCV was 0.047 and the mean MYO/C count density ratio was 5.4 +/- 2.42. Accuracy was 81% at a true MYO/C count density ratio of 6 in the phantom. It is concluded that the multi-slice SPET method using the left ventricular cavity VOI and a blood sample as a reference is a reproducible and accurate method for assessing myocardial [123I]MIBG uptake.