Evaluating castability as a property of dental casting alloys is an unworkable concept, since any measure is demonstrably affected by several external factors. Casting System Effectiveness (CSE) is shown experimentally and theoretically to depend primarily upon the time taken for the advancing front to freeze. A spiral tube mold was used to measure casting length obtained under variations in casting temperature; this pattern is proposed as a standard measure of CSE. A Bernoulli 'free-fall' numerical model is shown to reproduce the principal features of such casting, with some evidence of viscosity limitation of the turbulent flow at long casting lengths. Direct measurement of melt velocities confirmed turbulent flow in the mold. A method is also suggested for assessing the influence of investment porosity on CSE.