The paper is intended to show that the deficit in empirically based criteria for theory choice can be overcome by making use of a criterion of ontological plausibility. This move is not only defensible theoretically, but evident in the thinking of the scientific community since the seventeenth century. It is also intended to show how conceiving of science as based on iconic model building provides the resources needed to make judgments of ontological plausibility between competing theories even when these are referred to competing paradigms. The final step will be to suggest what must be done to sustain the concept of ontological plausibility when iconic modelling runs out.