An experimental investigation has been made of the application of the indicator-dilution bolus-injection technique to measure flow rate, mean transit time and vessel volume with steady flow in curved tubes and branching networks. Measurements were made with 131I in 2N KCl and 131I-human serum albumin in isotonic saline, plasma and whole blood under laminar flow conditions up to a Reynolds' number of 460. The flow rate estimations are more reliable than those obtained for flow through straight tubes, due to secondary flows tending to disperse the indicator uniformly over the vessel cross-section. Even so, estimations of mean transit time and vessel volumes, using accepted theoretical formulae, may be as much as 200% in error. Similar measurements have been made of the flow parameters with bifurcations and branching networks. The indicator is not partitioned between the branches in proportion to the flow rate, as is assumed in the original indicator dilution theory, except when there is complete symmetry of the flow. An alternative method of estimating the vessel volume was used and shown to give an accurate estimate of the true volume in these circumstances.