Information on other‐user interference is important in predicting the performance of high frequency (HF) systems. For many systems, especially frequency‐hopping systems, other‐user interference frequently is more important than man‐made noise from incidental radiators (for example, power lines) or atmospheric noise from lightning. Occupancy of individual channels and congestion of allocated HF bands (for example, fixed and mobile, amateur) vary with type of service, frequency, time of day, season, angle of arrival, type of receiving antenna, bandwidth, threshold, geographic location, and sunspot number. This paper summarizes channel occupancy and band congestion observations made in the continental United States and Europe in 1987 with SRI International's Mobile Spectrum Monitoring Unit. The data on band congestion are compared with the empirical model developed by researchers at the University of Manchester Institute for Science and Technology (UMIST), Manchester, United Kingdom. A concept for extending the UMIST model for congestion at a point to a model for congestion within a region is described.