When someone raises a concern in a relationship or workplace and walks away feeling like the accused, something specific has happened. It is not confusion. It is not misremembering. It is a sequence of moves, and understanding those moves as a structured pattern helps explain why confrontation so often produces the opposite of accountability. DARVO, first described by psychologist Jennifer Freyd in 1997, names that sequence: Deny, Attack, Reverse Victim and Offender.
What makes DARVO a strategic move
Game theory studies decision-making when outcomes depend not just on what one party does, but on what the other party does in response. Most people approach confrontation assuming the other person will either accept or reject what they raise. DARVO breaks that assumption by changing what game is being played.
The person who raised the concern entered a conversation about behavior. DARVO exits that conversation and starts a different one, a conversation about the accuser's credibility, motives, and emotional state. The original issue does not get resolved. It gets replaced.
This is not accidental. Each stage of DARVO serves a specific function in that shift.
The three stages and what each one does
Deny
The first move is flat denial. The behavior did not happen, or it happened differently, or the person raising the concern is misremembering it. This creates immediate friction. The target now has to argue for the existence of something they already know occurred, which pulls focus away from the behavior itself and redirects it toward whether memory and perception can be trusted.
Denial also resets the starting position of the conversation. Before DARVO, the starting position was: something happened and it caused harm. After denial, the starting position becomes: whether anything happened at all. Every point made from that position has to first clear the barrier of basic credibility.
Attack
Once denial is established, the next move is to shift attention to the person who raised the concern. Common attack patterns include questioning the person's motives, pointing to their past behavior, framing them as unstable or vindictive, or accusing them of raising the issue to gain something. The target of the original concern is now defending their character rather than the specific behavior they identified.
Research published in the Journal of Aggression, Maltreatment and Trauma found that when DARVO tactics are present, observers are more likely to question victim credibility and less likely to assign responsibility to the perpetrator. The attack stage is what produces that effect. It does not need to be aggressive to work. Quiet, persistent questioning of the other person's judgment or intentions produces the same result.
Reverse Victim and Offender
The final stage completes the role swap. The person who engaged in the harmful behavior now presents as the one who has been harmed. They were falsely accused. They are being treated unfairly. The attention they are receiving is itself a form of abuse.
This stage is where the strategic value of DARVO is clearest. In most social contexts, the person in the victim position receives support, benefit of the doubt, and reduced scrutiny. By occupying that position, the person using DARVO captures those advantages while the original target is left in the position of accused. The target now has to defend against the claim that they caused harm, which consumes the energy and attention that would otherwise focus on what they originally raised.
Why the target does not see it happening
DARVO works in part because it moves quickly and because each stage looks like a reasonable response in isolation. Denial looks like disagreement. Attack looks like self-defense. Role reversal looks like hurt feelings. A person experiencing all three in sequence, directed at them, under emotional stress, does not have the distance to recognize the pattern as a pattern.
Medical News Today notes that DARVO specifically exploits existing self-doubt. People who have already been conditioned to question their perceptions are more vulnerable to each stage. Denial lands harder when someone already doubts their memory. Attack lands harder when someone already questions their worth. Role reversal lands harder when someone already feels responsible for the emotional state of others.
The conditions that make someone a target in the first place, prior gaslighting, trauma bonding, patterns of self-blame, are the same conditions that make DARVO effective against them. The tactic is not random. It fits the vulnerability it encounters.
Observable behaviors at each stage
Because DARVO operates as a sequence, it produces specific observable behaviors. Knowing what to look for at each stage makes the pattern easier to identify in real time.
During the deny stage, watch for categorical statements like "that never happened" or "you're twisting what I said," refusal to engage with specific details, and redirection toward the target's emotional state rather than the facts of what occurred.
During the attack stage, watch for sudden introduction of the target's past behavior or character flaws, statements framing the concern as a personal attack, and accusations that the target has an ulterior motive for raising the issue.
During the reversal stage, watch for the person claiming emotional injury from the confrontation itself, positioning their distress as equivalent to or greater than what the target described, and appealing to observers or third parties for sympathy.
These behaviors do not always appear in strict sequence. Denial and attack often overlap. Reversal sometimes comes early, before denial is fully established. The sequence is a framework, not a script. What remains consistent is the direction: away from the original concern and toward the target's culpability.
What changes when the pattern is named
Naming DARVO as a pattern does two things. First, it gives the target something to orient around. When someone understands that what they are experiencing is a predictable sequence rather than a unique and confusing breakdown of communication, they gain some distance from the immediate emotional impact of each stage. The denial does not disappear, but it is less destabilizing when it is recognized as the first move in a known sequence.
Second, naming the pattern shifts the frame for observers. Research on DARVO consistently shows that people who understand how the tactic works are less likely to be moved by it. They are less likely to attribute credibility to the role reversal and more likely to hold attention on the original concern. Education about DARVO is one of the more direct tools available for reducing its effectiveness at the social level.
For clinicians working with clients who have experienced this pattern, the practical application is documentation. When a client can record specific exchanges close to the time they occur, the record provides a reference point that holds up against the rewriting that denial and reversal attempt to produce. Recognizing manipulation patterns in real time is a skill that develops over time and with support, not something a target is expected to arrive at on their own.
The difference between DARVO and ordinary conflict
Not every denial is DARVO. Not every counter-accusation is DARVO. The distinction lies in sequence, intent, and effect. Ordinary conflict involves two parties who each have a perspective on what happened and are working, however imperfectly, toward shared understanding. DARVO does not move toward shared understanding. It moves toward elimination of the original concern and reassignment of the victim role.
A useful test is to ask what the conversation is about after the response. If it is still about the original behavior, conflict is present. If it is now about the target's credibility, motives, or emotional stability, something else has happened. The conversation did not evolve. It was redirected. That redirection is DARVO in practice.
Clients who learn to ask that question, what is this conversation actually about now, gain a practical tool for orienting themselves during interactions that have historically left them disoriented. It does not stop the tactic from being used, but it interrupts the confusion the tactic depends on. Rebuilding after manipulation often begins with exactly this kind of reorientation toward observable evidence rather than the emotional weight of the moment.