While policy-makers understand the vast benefits of publicly funded and not-for-profit research, governance practitioners have the difficult task of defining processes that can best foster high performance for science and technology (S&T) labs not primarily driven by profits. This qualitative study develops a new taxonomy based on two dimensions, the nature of the funding relationship between the parent organisation and the S&T lab and the degree of interdependence between the lab's research units. We discuss the dynamics uncovered for each of the four archetypes with illustrative cases and argue that high-performance governance of S&T labs requires an internal coherence linking the processes of planning, funding and performance evaluation.