Over the last twenty years there has been a dramatic rise in subsidence claims for residential buildings. This has led to growing concern amongst insurance companies and the structural engineers who investigate these cases on their behalf. Their attention has begun to focus more on the engineering management of these cases. A review of the mechanism of subsidence damage to residential buildings and existing subsidence management procedures has shown that existing management procedures are often ad-hoc and uncoordinated, have been criticised for being inadequate, and have lead to solutions which have either been inappropriate or have failed to solve the problem. The review concludes that there is major scope for improving the decision-making process, by ensuring that it is based on a sound technical framework. This thesis describes an innovative approach to the engineering management of subsidence cases based on knowledge-based system (KBS) techniques. The Subsidence CAse Management System (SCAMS) is applicable to the multi-task domain and consists of three main components which address important facets of subsidence management. These are the diagnosis of a subsidence problem, the choice of an appropriate course of investigations, and the specification of effective remedial measures. SCAMS was developed using a KBS building shell- Kappa PC. The data structure adopted was object-oriented and incorporates production rules. The knowledge acquisition was based on multiple sources of knowledge using a variety of techniques. The system implementation was designed to result in a user-friendly system, and incorporates a highly graphical interface. Evaluation of the system was undertaken using test cases from industry. The system proved highly effective in all the tests and, in some cases, achieved more accurate results than the human expert. SCAMS represents a significant advance over existing approaches to the engmeenng management of subsidence cases. It provides many benefits to the construction industry including greater consistency in the evaluation of subsidence damage, improved quality of engineering judgement, reduced repair costs and specific guidance to -engineers for individual cases.