PURPOSE: To analyze Doppler waveform using branching vascular phantoms with stenosis. MATERIALS AND METHODS: Vascular phantoms consist of 12 different glasstubes with stenosis and side branch. Stenotic models of three different degrees were constructed to include severe (86%), moderate (64%), and mild (40%) stenosis calculated from internal diameters. There were two different branching type with or without distal reanastomosis and two different internal diameters of side branches. Pulsed Doppler waveforms were obtained at various sites of vascular phantoms. All hardcopy films were digitized and peak systolic velocity and acceleration times were obtained from the averages of the measurements in three consecutive triphasic peaks. RESULTS: In the simple stenosis model, peak systolic velocity at the prestenotic. site decreased when the stenosis was severe. The experiment using branching, phantoms without reanastomosis demonstrated that peak systolic velocities of main vessel at the prebranching site were higher than those of the simple stenosis model. In addition, peak systolic velocities of the branches wen; always higher than those of the main vessels except in phantom with mild stenosis and reanastomosing branch. In the latter phantoms, peak systolic velocities of the branches were lower than those of main vessels. Peak systolic velocities were always high in branches with smaller diameter. The measurements of acceleration time failed to demonstrate pulsus tardus in all cases probably because of collateral circulation and lack of elasticity of the glasstubes. CONCLUSION: This experimental study showed that degree of stenosis and type and size of side branch significantly affected Doppler waveforms in the vascular. phantoms. This result warrants that the effect of branch should be considered in the experimental vascular stenosis model. Further experimental study using an ideal vascular phantom with proper compliance is needed for clinical application.