The nature of smoking risk is first reviewed and a classification of procedures for assessing smoking behavior is presented. Areas requiring assessment include not only the traditionally measured smoking rate, but also the substance used and topography of consumption. Each of these areas may be assessed through a variety of self-report, observational, or indirect techniques. These techniques as well as some of their advantages and disadvantages are presented. Recently published (1975 to mid-1978) data-based smoking research appearing in four journals (Addictive Behaviors, Behavior Therapy, Journal of Applied Behavior Analysis, and Journal of Consulting and Clinical Psychology) is then reviewed with respect to measurement reliability and the use of multiple measures. Results show a strong tendency to assess only the risk area of smoking rate and a low frequency of appropriate measurement reliability checks, especially during baseline and treatment phases. Some of the implications of these results are discussed.