The extent to which a fast, nonequilibrium, and highly transient pore-scale process such as macropore flow can be predicted is very often debated, although little research has been conducted to investigate this issue. The validity of approaches to “upscaling” transport predictions from pore through Darcy to landscape scales critically depends on the answer to this question. We developed a simple conceptual model of soil susceptibility to macropore flow, based on a synthesis of existing experimental information. The conceptual model takes the form of a decision tree, which classifies soil horizons into one of four susceptibility classes on the basis of easily available site and soil factors. The model was tested against an independent database of tracer breakthrough experiments on undisturbed soil columns collated from the literature (n = 52), using the pore volumes drained at peak solute concentration, tp, as a measure of the strength of macropore flow. Analysis of variance for tp as a function of susceptibility class showed that the overall model was significant. A significant proportion of the residual variation in tp could be attributed to variation in clay content within one of the susceptibility classes. Some important sources of experimental error were also identified that may account for much of the remaining unexplained variation. We concluded that macropore flow is predictable to a sufficient degree from easily available soil properties and site factors. The simple classification tree developed in this study could be used to support hydropedological approaches to quantifying the spatial distribution of contaminant leaching at the landscape scale by providing the basis for class pedotransfer functions to estimate model parameters related to macropore flow. Such an approach has been implemented in the European project FOOTPRINT.