Do Mental Health Apps Actually Work? What the Research Says

The Short Answer

Yes, on average, but with real caveats. A 2024 meta-analysis published in World Psychiatry, pooling 176 randomized controlled trials, found that standalone smartphone mental health apps produced small but statistically significant improvements in depression symptoms (effect size g = 0.28) and generalized anxiety (g = 0.26), compared with inactive controls.

The Effect Size Is Small, and That Matters

"Statistically significant" doesn't mean "life-changing." A small effect size means the average improvement is real but modest -- helpful for many people with mild-to-moderate symptoms, but not comparable to intensive therapy or medication for more severe conditions. The same review found depression-specific, CBT-based apps performed better (g = 0.38) than general anxiety apps (g = 0.20), so the specific app and technique matter.

The "Digital Placebo" Caveat

A notable finding in the same analysis: apps looked considerably more effective when compared against an inactive control (doing nothing) than against an active comparison (another treatment or even a different app). That gap suggests part of the measured benefit may come from simply engaging with something structured and paying attention to your mental state, not necessarily from the specific technique inside the app.

Nearly Half Use CBT for a Reason

The same review noted that close to half of the apps studied were built on cognitive behavioral therapy frameworks, which had the strongest evidence base among the approaches tested -- a reasonable filter if you're choosing between apps and want to weight the odds toward something with actual research behind it.

What This Means Practically

Apps are a reasonable, evidence-backed option for mild-to-moderate anxiety and low mood, particularly CBT-based ones, used consistently rather than downloaded and abandoned. They are not shown to be a substitute for therapy in moderate-to-severe cases -- see the app-vs-therapy question addressed separately for that distinction.