Habit Reversal Training for BFRBs: What the Evidence Actually Shows

Habit reversal was introduced by Nathan Azrin and Gregory Nunn in 1973 as a treatment for nervous habits and tics. In their first report they treated 12 people whose habits ranged from nail biting to eyelash picking in a single session, and the habits were virtually eliminated on the first day for everyone who followed the instructions.

Modern practice keeps three components at the core: awareness training (learning to catch each occurrence and the urge that precedes it), competing response training (doing a physically incompatible action instead), and social support. The TLC Foundation for BFRBs describes habit reversal as an early treatment developed by Azrin and Nunn that has been examined more than any other in research studies.

Habit reversal now often travels inside larger packages, such as the comprehensive behavioral (ComB) model, which adds work on the sensory, cognitive, and situational triggers around a behavior. The shared engine is still awareness plus a competing response.

Trichotillomania (hair pulling disorder) is the BFRB with the deepest trial record, and three meta-analyses point the same way. A 2007 review of seven trials found habit reversal superior to medication, with an effect size of 1.14 against 0.68 for clomipramine and 0.02 for SSRIs, meaning the SSRIs barely beat placebo.

A 2014 meta-analysis of 11 randomized trials reported a large pooled effect of 1.41 for behavior therapy, compared with a moderate 0.41 for serotonin reuptake inhibitors. A 2020 update covering 24 trials and 857 participants found a large benefit again for behavior therapy with habit reversal components, a standardized mean difference of 1.22 against control conditions.

The authors of the 2020 review put it plainly: habit-reversal-based behavior therapy has demonstrated the largest treatment effects and the strongest evidence base for reducing hair-pulling symptoms.

The bars below are effect sizes from two meta-analyses, shown as magnitudes (larger means a bigger benefit). Within each analysis the comparison is apples to apples, and behavior therapy outsizes the medications every time. SSRIs, at 0.02, are essentially flat.

Habit reversal is also the engine inside the leading behavioral treatment for tics. In a 2010 randomized trial of 126 children with Tourette or chronic tic disorder, a habit-reversal-based program produced a response in 52.5 percent of children against 18.5 percent in a control condition, and 87 percent of responders still held their gains six months later.

For nail biting the picture is more limited. A dermatology review notes that stimulus control, habit reversal, and medication are used to manage onychophagia, alone or more often in combination, while calling nail biting a difficult behavior to change. Randomized trials specific to nail biting are older and small, so the strong hair-pulling and tic results are the firmest ground.

That is the honest frame for anyone reaching for these numbers. Habit reversal is a better-established behavioral approach for BFRBs as a family, the trial evidence is deepest for hair pulling and tics, and applying it to nail biting is reasonable extrapolation rather than a settled, nail-biting-specific result.

Effect sizes are shown as magnitudes. Different reviews report the sign differently, so read the note column, and remember that a larger number means a larger measured benefit.

The effect sizes are genuinely large, and they also carry limits worth stating out loud.

Every figure on this page traces to a randomized trial or a randomized-trial meta-analysis, linked in the sources list, and was confirmed against a supporting quote from that source.

Effect sizes are reported as magnitudes for readability. Some reviews report negative standardized mean differences, where a negative value means symptoms went down; the size of the number is what matters for comparison.

The strongest evidence base is in trichotillomania (hair pulling) and tics. Nail-biting-specific randomized trials are older and small, so we frame habit reversal as well-established for BFRBs broadly rather than proven for nail biting in particular.

Comparisons across different meta-analyses are not exact, because samples and control conditions differ. The fair comparison of therapy against medication is the one made inside a single analysis, which is how the chart is built.

We omit any claim that habit reversal cures a habit or works for a fixed share of people, and we omit a single effect size for nail biting, because those could not be verified against a primary source.