In the theoretical analysis of arterial networks the existence of geometric optimisation has long been suggested, although observational studies have not yet fully corroborated these theories. Since this could be due to experimental flaws, the aim of this study was to establish the validity of arterial geometric optimisation using a new experimental design and to assess the influence of ageing. Solid vascular casts of arterial mesenteric branching systems of 8-wk-old and adult dogs (beagles) were used to examine vascular diameters and branching angles, the latter in a manner that allowed optimisation of the line of view, thus minimising distortion errors due to a line of view not normal to the branching plane. Internal and external vessel diameters were found to be in accordance with the theoretical principle of minimum work (8-wk-old internal: r = 0.994; adult internal: r = 0.971; adult external: r = 0.985). Although branching angles were found to be in agreement with basic qualitative principles of arterial branching geometry, the measurements still showed a large amount of scatter and were generally smaller than expected on theoretical grounds, despite the newly designed measuring technique. These branching angles demonstrated small age-related differences. However, when biological cost was considered per bifurcation, surprisingly, guidance towards minimum lumen volume and pumping power with increase in age could clearly be demonstrated (P<0.001). It is concluded that our findings support the existence of a degree of arterial geometric optimisation in favour of minimum lumen volume and pumping power, increasing with age. Future investigations should focus on the biofeedback mechanisms involved.