The best current estimate is that roughly 16–25% of people who survive a suicide attempt will make another nonfatal attempt within five years, depending on setting and history. Most people do not repeat. A 2022 meta-analysis in Molecular Psychiatry (Liu and colleagues) put the 5-year rate of repeat nonfatal attempts at about 16–17% in the general clinical population, while large health-system cohorts have reported figures closer to 20–25% by five years. Read that range as context for prevention, not as a personal forecast.

The variation isn't noise. It tracks who was counted, whether they had prior attempts, what care they received after the first, and whether researchers followed a hospital cohort, an insured population, or a community sample. Below I unpack the numbers behind the range, the study designs that produce them, and the practical levers that shift the odds.

What the Best Meta-Analytic Estimate Says About 5-Year Recurrence

Meta-analyses pool many cohorts, so they're the strongest single source for this question. Liu's team analyzed 16 prior reviews covering well over 300,000 people and reported a 0.5-year rate of repeat nonfatal attempt near 16%, a 1-year rate around 19%, and a 5-year rate not much higher at roughly 16–17% across their pooled estimates. That flatness is the striking finding: most repeat attempts happen early, and the cumulative rate plateaus rather than climbing steadily.

The same analysis found a 5-year rate of repeat nonfatal self-harm around 19% and a 5-year rate of actual deaths by suicide around 1.6%. Those three numbers belong together. Repeat self-harm is common; death is rare. Conflating them is the biggest error readers make.

One important caveat: the meta-analysis's update window means later primary studies aren't fully captured in every pooled estimate. If you see a 2024 or 2025 primary cohort reporting a higher 5-year figure, that's not a contradiction — it's a different population measured with better record linkage.

The practical takeaway from the pooled data is that a survivor's highest risk window is the first 6 to 12 months, and the annualized risk in years 2 through 5 is lower than most people assume.

Why Health-System Cohorts Report 20–25% by Five Years

When researchers follow a defined insured or hospital population with electronic records, the 5-year repeat attempt rate tends to run higher than pooled meta-analytic figures. A large cohort study of patients treated for deliberate self-harm in a regional health system reported approximately 20% making a repeat attempt within five years. Other HMO-type cohorts have landed in the same neighborhood.

The reason is selection. A hospital cohort includes only people who presented for care, which skews toward higher intent and more prior attempts. Insured populations have continuous records, so a repeat attempt that never reached an emergency department but was recorded at an urgent care visit still counts.

Consider a worked implication. If a health system with 100,000 annual self-harm presentations wants to size its aftercare program, a 20% five-year repeat rate projects 20,000 repeat attempts over five years — about 4,000 per year — from that starting group alone. That arithmetic is why follow-up contact programs get funded. It's also why a program that reduces the repeat rate by even 3 percentage points would prevent roughly 600 repeat attempts per year in that system.

National register studies in countries with linked health data (Denmark, Sweden, and Taiwan have the strongest) largely confirm the 15–25% band, with the high end concentrated among people with multiple prior attempts.

The 16–25% Range Breaks Down by How You Count

The single biggest driver of the reported percentage is what counts as a repeat attempt. Methods that capture only medically treated events produce lower numbers than methods that include self-reports. The table below shows how the same underlying population can produce different published figures.

Counting method What gets captured Reported 5-year repeat rate
Emergency department only Visits coded as self-harm or suicide attempt 10–15%
Hospital + outpatient records Adds urgent care, psychiatric intake 18–25%
Record linkage + structured interview Adds self-reported attempts not coded 25–30%
Meta-analysis, pooled across designs Weighted blend of the above 16–17%

Read the column, not just the number. A figure of "16%" from a meta-analysis and a figure of "25%" from a cohort study can both be correct if the counting rules differ. Studies in the United States also tend to miss attempts treated at facilities outside a patient's network, which pushes published U.S. rates downward relative to Nordic registers.

For anyone reading a specific study, check three things: the definition of "attempt," the observation window, and whether deaths were censored or counted separately. Those three decisions explain most of the spread you'll encounter.

Prior Attempts Change the Five-Year Odds Sharply

A first-ever attempt and a third attempt are not the same clinical event. Research consistently shows that prior attempts are among the strongest predictors of future attempts, and the five-year repeat rate climbs with each prior attempt. In pooled data, people with one prior attempt show a repeat rate in the high teens to low twenties; people with multiple prior attempts push toward 30% or more within five years, and some high-risk clinical samples report 40% within two years.

That gradient is why risk screening tools ask about history. A validated tool like the Columbia Suicide Severity Rating Scale captures prior attempts as a core item. If a clinician knows a patient has made two prior attempts, the conversation about follow-up intensity changes.

Here's the counterintuitive part. Most people with a first attempt do not make a second. A first-time attempt in an insured population might carry a five-year repeat risk near 15%, meaning roughly 85 in 100 patients don't return with a repeat attempt. Treating every first-time presenter as high-risk with aggressive interventions has a cost in civil liberties and stigma, which is where the judgment call sits.

The practical distinction: focus the most intensive contact on people with prior attempts, a recent attempt, and co-occurring substance use or severe depression, because those factors compound.

The Highest-Risk Window Is the First Six Months

Timing matters as much as the percentage. Across cohort studies, the month immediately after discharge is the most dangerous period, with the first week to 90 days carrying risk many times higher than average. Repeat attempt rates are front-loaded: roughly half of the repeat attempts in a five-year window occur within the first year.

This has direct implications for aftercare design. A program that makes one follow-up call at 90 days misses the peak. A program with contact at 24 hours, 72 hours, one week, and one month covers more of the risk window with fewer total contacts than a spread-out schedule.

Australia's hospital-based "SUPPORT" type programs and the U.S. "Caring Contacts" approach both front-load contact this way. Caring Contacts, developed by Jerome Motto in the 1970s and refined by Gregory Carter's team in Australia, sends brief non-demanding messages — letters, postcards, texts — at intervals starting within days of discharge.

For a reader looking at the percentage question, the timing point is the actionable one. The five-year number is a planning figure. The early-months number is where interventions change outcomes.

What Follow-Up Contact Programs Do to the Repeat Rate

The most-studied intervention for reducing repeat attempts is low-burden follow-up contact. A meta-analysis by Milner and colleagues in Psychological Medicine found that brief contact interventions reduced repeat suicide attempts by roughly 15–20% on a risk ratio basis, though the authors cautioned that evidence quality varied. More intensive treatments — dialectical behavior therapy and cognitive behavioral therapy for suicide prevention — show larger reductions in randomized trials, with relative risk reductions in the 40–50% range for repeat attempts at roughly one to two years.

Translate that to the percentage question. If a cohort's baseline five-year repeat rate is 20%, a contact program with a modest 15% relative reduction projects a rate near 17%. A structured psychotherapy program with a 45% relative reduction projects near 11%.

Intervention Relative reduction in repeat attempts Basis
Brief contact (postcards, texts, calls) ~15–20% Meta-analysis, Milner et al., Psychological Medicine
DBT for high-risk adults ~40–50% at 1–2 years Multiple randomized trials
CBT for suicide prevention ~40–50% at 1–2 years Randomized trials
Safety planning + means restriction Reduces attempts and deaths Cochrane reviews

Means restriction is the intervention most consistently associated with reduced suicide deaths. Firearm storage counseling at discharge is now a standard item in U.S. hospital discharge protocols for patients with suicide-related admissions, and it belongs in the same conversation as the percentage figures.

Note the honest limitation: most trials measure 12 to 24 months, not five years. The five-year projections above are arithmetic extensions, not directly measured outcomes, and should be read that way.

Age, Sex, and Method Shift the Odds in Predictable Directions

Age and demographic patterns matter. Younger people, particularly adolescents and young adults, show higher rates of repeat nonfatal attempts relative to completed suicide. Older adults attempt less often but die more often per attempt, because they choose more lethal methods and are more likely to live alone. The National Violent Death Reporting System in the U.S. shows that older men have markedly higher case fatality from attempts than younger women.

Method choice matters too. Attempts by ingestion (overdose) are the most common and the least lethal, and they account for a large share of repeat attempts. Attempts by firearm carry a case fatality rate near 90%, which means they're underrepresented in repeat-attempt statistics simply because fewer people survive to be counted.

Sex differences show a similar pattern. Women make more nonfatal attempts; men die by suicide at roughly three to four times the rate. That asymmetry is the "gender paradox" in suicide research, and it's why repeat-attempt percentages and suicide death statistics appear inconsistent when viewed side by side.

If you're applying these numbers to a specific population, adjust for age, sex, and method — the general 16–25% range can be too low for a high-risk acute inpatient sample and too high for a low-lethality community sample.

Screening and Risk Tools and their Limits for This Number

Tools exist to estimate individual risk, but none of them predicts a specific person's next attempt with reliability. The Columbia Suicide Severity Rating Scale categorizes current and past suicidality and is widely used in emergency departments. The Safety Planning Intervention is a structured tool for producing a written coping plan. Neither outputs a personal probability, which is exactly the point: the five-year percentage is a population figure, not a test result.

Screening also carries risks. False positives can trigger involuntary holds and erode trust; false negatives give a false sense of safety. A 2024 systematic review found that the predictive accuracy of suicide risk scales in emergency settings is modest, with area-under-curve values typically in the 0.6 to 0.7 range — better than chance, not a diagnostic.

The practical guidance: use screening to structure the conversation, not to decide who gets care. Ask directly about prior attempts, current plans, access to lethal means, and substance use. Direct questions do not increase suicide risk — that myth has been tested and rejected.

For families, that means asking your loved one directly. "Are you thinking about killing yourself?" is a legitimate question, and asking doesn't plant the idea.

Recovering Well After an Attempt: What Changes the Long-Term Picture

Most people who survive an attempt go on to live full lives, and the five-year percentage should be read in that light. Longitudinal research following people after suicide attempt shows that with treatment and support, the majority of people with a first attempt — around 75 to 85% — do not go on to die by suicide, and a large majority avoid another attempt. This is not a guarantee, but it's the trend worth holding onto.

The recovery factors that show up consistently are: treating the underlying condition (depression, substance use disorder, bipolar disorder), engaging with psychotherapy, reducing access to lethal means during high-risk periods, and having at least one person to call. Social connection appears repeatedly in large studies as a protective factor, from the U.S. National Violent Death Reporting System to international cohort data.

One edge case worth knowing: people who stop seeking help after an attempt — because of shame, cost, or lack of access — are not counted in many follow-up studies. Researchers call this attrition bias, and it means published repeat rates may underestimate what happens to people who disengage from care entirely.

If you're supporting someone after a suicide attempt, the most practical thing you can do is maintain contact. Data on follow-up programs show that a brief message helps. The five-year number matters less than whether someone is around to send that message in the first week.

FAQ

Is the repeat attempt rate higher after a first attempt or after a second?

After a second. People with one prior attempt show five-year repeat rates in the high teens to low twenties; people with multiple prior attempts push toward 30% or more, with some high-risk clinical samples reporting 40% within two years. Each prior attempt raises the baseline of future risk, which is why history is a core screening item.

Do most people who attempt suicide attempt again?

No. Roughly 75 to 85% of people who survive a first attempt do not make another nonfatal attempt within five years. The 16–25% figure describes the minority who do, and it varies by population. This is why the recovery literature emphasizes support figures over recurrence figures.

How does the five-year repeat rate compare with the five-year suicide death rate?

Much lower for death. Pooled data in the 2022 Molecular Psychiatry meta-analysis put the 5-year suicide death rate near 1.6%, compared with 16–17% for repeat nonfatal attempts. The death rate is approximately one-tenth the repeat-attempt rate. Both are used separately in public health planning.

Does a nonfatal attempt increase future risk in the same way across countries?

No. Method availability dominates. Nordic register studies and U.S. cohorts report broadly similar repeat-attempt bands (15–25%), but method-specific shares differ sharply because of firearm availability, which shifts the ratio of attempts to deaths. In 2017, more than half of U.S. suicide deaths involved firearms, compared with a much lower share in Denmark or Taiwan.

Can a safety plan reduce the chance of another attempt?

Yes, probably. Safety planning added to usual care is associated with reduced suicidal behavior in randomized trials, and Cochrane reviews link it with fewer repeat attempts. The evidence is strongest for structured safety planning combined with means restriction counseling at discharge rather than either alone.

What should a family member do in the first week after a hospital discharge?

Maintain contact, ask directly about suicidal thoughts, and help secure lethal means at home — firearms stored off-site, medications locked. The first 90 days post-discharge carry the highest risk. Offering practical presence (rides to appointments, sitting with them) often helps more than advice.

Does asking someone about suicide raise their risk of attempting?

No. Research has tested this repeatedly and consistently finds that asking about suicidal thoughts does not increase risk. Screening conversations can actually make it easier for someone to disclose distress. The belief that asking plants the idea is a common myth worth correcting.

Are repeat attempts in young people more common than in older adults?

Yes for nonfatal repeat attempts. Adolescents and young adults show higher repeat-attempt rates relative to completed suicide, while older adults attempt less often but die more often per attempt. National Violent Death Reporting System data show older men in the U.S. have a higher case fatality from attempts than younger women, which reverses the pattern of who appears in the repeat-attempt statistics.

This article is general educational information and is not a diagnosis, treatment plan, or clinical recommendation. Suicide risk is individualized, and no percentage can predict what will happen to a specific person; decisions about care belong with a qualified mental health or medical professional who knows the situation. If you or someone you know is in immediate danger, contact local emergency services or a crisis line in your country (for example, 988 in the U.S. and Canada) right away.

Resource Library

About This Article

This guide was produced by Cerebral Sisters as part of our mission to provide accessible, evidence-informed mental health education. It is reviewed against current clinical guidelines and research literature, and is intended for educational purposes only.