Priming tasks are frequently used reaction time (RT) paradigms to study automatic panic-related associations. Such tasks involve the (brief) presentation of a prime (e.g., a word), followed by a target requiring a response (e.g., categorization). The RT needed to respond to the target serves as an index of the ‘associative match’, i.e., the prime can either facilitate or aggravate reactions and thereby decrease or increase RTs. To illustrate, the priming study by Schniering and Rapee (1997) included associatively related and associatively unrelated prime–target pairs, which participants had to categorize as ‘words’ or ‘non-words’. Most relevant were trials where primes referred to bodily sensations and targets to catastrophic outcomes (e.g., breathlessness-suffocate; dizzy-faint). Contrary to predictions of PD models, there was no difference in RTs between panic patients and controls on panic trials (see also Schneider and Schulte, 2008 and McNally et al., 1997). To the best of our knowledge, the study by Hermans et al. (2010) is the first to show the expected pattern, i.e., a faster RTs for panic trials in panic patients than in controls.