A logical model of the process of calibrating a new code of practice has been developed. The hierarchical system of statements was derived from an initial statement that structures built in the future should be no less reliable, on average, than existing structures. The other statements which are connected by fuzzy logical operators and are necessary conditions on the first statement, cover random hazards, human controls on tolerance and quality, human systems such as design and construction and the process of code calibration itself. Data were acquired for input to the model from two individuals who were involved in the calibration of BS 5400 against BS 153, the new and old codes for steel bridge design in the U.K. These data were intended only for use in illustrating the potential of the fuzzy logic modelling. However, the analysis showed that both donors felt that statements outside the detailed calibration process were less dependable than those inside, particularly those concerned with the human activities of design and construction. Within the calibration process the quality of data and the loading model was considered to be the least dependable part of the analysis. The fuzzy logic method described in this paper, with some development of consistent ways of obtaining subjective estimates from various sources, could be used to report a consensus on this type of complex process.