Patent data are an outstanding resource for the study of technical change. Alongside other science and technology (S&T) indicators such as R&D expenditure and personnel or innovation-survey data, patent data provide a uniquely detailed source of information on inventive activity and the multiple dimensions of the inventive process (e.g. geographical location, technical and institutional origin, individuals and networks). Furthermore, patent data form a consistent basis for comparisons across time and across countries. Yet such data are complex, and patent-based indicators must be designed and interpreted carefully. Since the publication of the first OECD manual on patents in 1994 (The Measurement of Scientific and Technological activities: Using Patent Data as Science and Technology Indicators), significant progress has been made both in data provision and statistical analysis based on patent data. The 2009 edition of the OECD Patent Statistics Manual takes stock of the recent developments in the field. It provides guiding principles for the use of patent data in the context of S&T measurement, and recommendations for the compilation and interpretation of patent indicators in this context. It aims to show what patent statistics can be used for, what they cannot be used for, and how to count patents in order to maximise information on S&T activities while minimising statistical noise and biases. Finally, it describes how patent data can be used in the analysis of a wide array of topics related to technical change and patenting activity including industry-science linkages, patenting strategies by companies, internationalisation of research, and indicators on the value of patents.