A theoretical description of the relationship between the count rate detected in the cardiac region of interest and the arterial organ input function in dynamic scintigraphy is presented. It is shown that, provided the time-activity curve for the heart is corrected for extra-cardiac and unlabelled activity, it is proportional to the arterial organ input over both the first-pass and equilibrium phases of the passage of an intravenously injected radiopharmaceutical. A practical example demonstrating the value and validity of correcting the cardiac curve in dynamic radio colloid scintigraphy is described. Estimates of the colloid clearance rate to reticuloendothelial sites using the cardiac curve without correction were significantly lower (mean 0.085/min) than those derived from the liver uptake curve (mean 0.213/min). However, corrected cardiac curve clearance rates (mean 0.225/min) were not significantly different from the liver ones. Also, the corrected cardiac curve clearance values correlated linearly with liver curve values (correlation coefficient 0.89, standard error of the estimate 0.021/min), whereas the uncorrected values showed no significant correlation. Thus, correction of the cardiac curve gave clearance rate values that were both more accurate and more precise than those obtained without correction.