Self-monitoring has become a popular assessment tool in treatment outcome research and descriptive psychopathology for panic and anxiety disorders. Commonly used formats of self-monitoring for each anxiety disorder are described. Self-monitoring is compared with self-report estimation. On the one hand, self-monitoring may be influenced by reactivity, as well as by attentional and judgmental biases that underlie fear and anxiety. On the other hand, self-monitoring may minimize certain influences that tarnish estimation, such as response demand biases, memory deficits, inflation of aversive experiences, and availability heuristics. Issues for statistical analysis of self-monitored data for panic and the anxiety disorders are outlined. Finally, directions for future research and clinical use of self-monitoring are discussed.