This paper provides empirically-based explanations for the observed low, and partial, use of information systems development (ISD) methodologies. The research reported in this paper examines the process of methodology selection and use in an intensive field study of the early stages of systems development and notes ongoing adaptation of methodologies in response to changing contingencies. The paper concludes that, in many circumstances, methodology adaptation is a necessary part of successful systems development. It also suggests that deeper understanding of the attitudes of different stakeholders to selecting and using ISD methodologies may help to explain this adaptation.