BackgroundFear conditioning (FC) in rodents is the most used animal model to investigate the neurobiology of posttraumatic stress disorder (PTSD). Although research using FC has generated a better understanding of fear memories, studies often rely on mild or moderate FC training and behavioral analysis generally focuses on measuring freezing responses within few test sessions. New methodWe introduce the M-Maze task, a system that measures extinction of conditioned fear using suppression of operant behavior. The apparatus consists of an M-shaped maze where rats are trained to alternate nose poking at two pellet dispensers. Proximity sensors measure the animal’s locomotion, as well as the latencies and number of operant behaviors. Here we also describe the protracted aversive conditioning (PAC), a rat model of severe fear that induces resistant extinction following a 4-day conditioning protocol that combines delay, unpredictable, and short- and long-trace conditioning. ResultsAn intense one-day auditory FC protocol induced a sharp elevation in transit time and suppression of nose pokes by conditioned cues, but in contrast to what is found in PTSD patients, fear extinction was rapidly observed. On the other hand, PAC alone or in combination with exposure to single prolonged stress induced persistent extinction impairments in M-Maze tests, as well as enhanced anxiety, and social withdrawal. Comparison with other existing methodsThe M-Maze task is fully automated and allows multiple animals to be tested simultaneously in long-term experiments. Moreover, PAC training can be an alternative approach to study extinction-resistant fear. ConclusionsThe M-Maze task allows rapid and unbiased measurements of fear-induced suppression. We suggest that long-term assessment of extinction impairments would lead to a better understanding of the neurobiology of persistent fear and the screening for new therapies.