When a conditioned response stops occurring, psychologists call it extinction. This happens when the event that originally reinforced the behavior is removed. In classical conditioning, extinction occurs when a signal is repeatedly presented without the outcome it was paired with. In operant conditioning, it occurs when a voluntary action no longer produces the reward or consequence that maintained it. Either way, the behavior weakens and eventually stops, though “stops” turns out to be more complicated than it sounds.
How Extinction Works in Classical Conditioning
The classic example comes from Pavlov’s dogs. A bell is paired with food until the bell alone triggers salivation. If the bell keeps ringing but food never arrives, the dog gradually salivates less and less until the response disappears. The conditioned stimulus (the bell) loses its power because it no longer predicts the unconditioned stimulus (the food).
This same process applies to learned emotional responses. If you once got food poisoning at a particular restaurant, walking past it might make you feel queasy. But if you pass that restaurant dozens of times without getting sick, the queasy feeling fades. The association between the place and the illness weakens through repeated exposure without the negative outcome.
Extinction in Operant Conditioning
Operant extinction follows a similar logic but involves voluntary behavior. A child who throws tantrums to get screen time will eventually stop if tantrums never produce screen time again. A rat pressing a lever for food pellets will stop pressing if the pellets stop coming. The behavior was maintained by its consequences, and when those consequences disappear, so does the behavior.
Before the behavior fades, though, something counterintuitive happens. The behavior often gets worse first. This temporary spike is called an extinction burst: the frequency, intensity, or duration of the behavior increases before it decreases. A child whose tantrums used to work may scream louder, try new tactics, or persist longer before finally giving up. This burst is a predictable part of the process, not a sign that extinction is failing. Parents and caregivers who give in during the burst accidentally teach the child that escalation works, which makes the behavior harder to eliminate next time.
Why Some Behaviors Are Harder to Extinguish
Not all conditioned responses fade at the same rate. One of the strongest findings in behavioral psychology is the partial reinforcement extinction effect. Behaviors that were rewarded every single time are actually easier to extinguish than behaviors that were rewarded inconsistently. This seems backward, but the logic is straightforward: if you’ve only ever been rewarded every time, the absence of reward is immediately noticeable. If you’ve been rewarded unpredictably, you’re used to going without, so it takes longer to recognize that the reward is truly gone.
Several explanations account for this. One is that inconsistent reinforcement makes extinction feel more like normal conditions. You’re accustomed to stretches without a reward, so a few more unrewarded attempts don’t register as a meaningful change. Another explanation focuses on the math of detection: when rewards were already infrequent, you need more unrewarded trials before you can statistically detect that conditions have changed. This is why slot machines, which pay out on a highly unpredictable schedule, produce behavior that is notoriously resistant to extinction. The gambler keeps pulling the lever long after the odds have turned against them, because intermittent reward trained them to persist through dry spells.
What Happens in the Brain During Extinction
Extinction requires coordinated changes in two key brain areas: the prefrontal cortex (the region behind your forehead involved in decision-making and emotional regulation) and the amygdala (a deeper structure that processes fear and emotional memories). During fear extinction specifically, the prefrontal cortex activates and suppresses the amygdala’s output. This reduces the fear response and allows extinction to proceed.
When the prefrontal cortex is damaged or suppressed, the amygdala’s emotional associations go unchecked, and conditioned fear responses persist longer than they normally would. This circuit helps explain why chronic stress, which impairs prefrontal cortex function, can make it harder to let go of learned fears.
One of the more important discoveries in extinction research is that extinction does not simply erase the original association. For years, scientists debated whether extinction represented genuine “unlearning” or whether the brain was forming a new memory that suppressed the old one. The current understanding is that neither description is entirely right. The original association weakens, but most of the neural connections involved don’t return to their pre-learning state. The old memory remains partially intact, which explains why extinguished responses can come back under certain conditions.
Why Extinguished Responses Come Back
If extinction truly erased the original learning, the conditioned response would be gone permanently. Instead, it can resurface in at least three well-documented ways.
Spontaneous Recovery
After a conditioned response has been extinguished, it can reappear simply with the passage of time. If you extinguish a fear response on Monday, the fear may partially return when you encounter the same stimulus on Friday. The size of this recovery is proportional to how strongly the response was suppressed: the more extinction training that occurred, the larger the spontaneous recovery can be. This doesn’t mean extinction failed. Each time spontaneous recovery occurs and the response goes unreinforced again, it weakens further.
Renewal
Extinction learning is surprisingly tied to the environment where it happened. If a fear is conditioned in one setting (say, a particular room) and extinguished in a different setting, the fear returns when you go back to the original room. This is called the renewal effect, and it reveals that extinction is partly context-dependent learning. Even moving to a completely new, neutral environment can trigger renewal, which suggests that simply leaving the extinction context is enough to bring the old response back.
This has real implications for therapy. A person who overcomes a phobia in a therapist’s office may find the fear creeping back in the environment where it was originally learned. Effective treatment accounts for this by practicing extinction across multiple settings.
Reinstatement
Even after successful extinction, a few unexpected encounters with the original reinforcer can restore the conditioned response. In fear conditioning experiments, presenting the aversive event a handful of times after extinction causes the extinguished fear to return, but only when the re-exposure and the test happen in the same environment. In one study, participants who experienced all phases of the experiment in the same setting showed clear fear recovery after just four presentations of the aversive stimulus. Participants who experienced those presentations in a different setting did not. The reinstated fear then re-extinguishes with further unreinforced exposure, so it’s not a permanent setback.
Extinction in Therapy
Exposure therapy is the most direct clinical application of extinction. By repeatedly confronting a feared stimulus without the expected negative outcome, the fear response weakens. For specific phobias like fear of spiders, heights, or flying, exposure therapy helps over 90% of people who complete the full course of treatment. It’s often the only therapy needed for a specific phobia.
For more complex conditions like PTSD or OCD, exposure therapy is typically one component of a broader treatment plan. The principles are the same: controlled, repeated exposure to the trigger without the feared outcome, allowing the brain’s prefrontal cortex to build new associations that suppress the old fear response. The challenge with these conditions is that the extinction process can be slower, and the phenomena of spontaneous recovery, renewal, and reinstatement mean that maintaining progress requires ongoing practice in varied real-world contexts rather than a single therapeutic setting.