A computer-based, preventive conservation decision support model is being developed for museum professionals. Its main applications include building surveys and planning exercises where it is necessary to make choices from a wide range of design features that control nine agents that cause material damage in museum collections. The basic components of the decision support model include an audit of building features and a collections damage model. The audit assesses control features and generates sequences of control feature improvements derived from expert judgement. The collections damage model ranks control measures by accounting for the nature and susceptibility of the collections to each agent of deterioration. The preservation feature audit and the method by which expert judgement is incorporated into the audit and collections damage model is described in this paper.