It is argued that technological forecasting is necessary but by itself insufficient for overall effective R & D planning, because which of the many available technological alternatives are exploited will be a function of their social efficiency. Social forecasting is the necessary complement to technological forecasting for effective R & D planning. Some of the problems of social forecasting are explored in a comparison with technological forecasting. Techniques of social forecasting in ascending complexity are discussed together with their relative advantages, limitations, and requirements. An example of the problems and some possible approaches to social forecasting is given, using the case of the impact of social change on the food and chemical industry and domestic purchases. Finally a recommended procedure for social forecasting is described.