How problem gambling is diagnosed: DSM-5, PGSI and SOGS
Gambling addiction diagnosis is made by a clinician using DSM-5 or ICD-11. Screening questionnaires such as the PGSI and SOGS estimate risk, but they do not replace a clinical assessment.
Screening and diagnosis are not the same thing
A gambling screen is a short questionnaire. It asks about behaviour, harm and loss of control, then gives a score or a simple risk result. That result can help someone decide whether to speak with a professional, use support tools or read more about signs of a gambling problem.
A screen does not decide whether a person has a mental disorder. It cannot weigh context, rule out other explanations or assess safety in the way a trained clinician can.
Screening also has a different purpose from diagnosis. Public health bodies use screens to estimate risk in groups of people. Treatment services may use them at triage. Individuals may use them as a prompt to reflect on gambling patterns.
None of those uses is the same as a clinical judgement. A diagnosis requires an interview, evidence of distress or impairment, and attention to timing, mood, substances and other possible explanations.
Psychiatrists, clinical psychologists and addiction specialists can assess gambling disorder, depending on the health system and service. They look at gambling history, money pressure, relationships, work, mental health and immediate risks. For prevention tools before problems escalate, see the responsible gambling guide.
The DSM-5 criteria for gambling disorder
DSM-5 is a diagnostic manual, not a self-test. Its gambling disorder criteria focus on a repeated pattern of gambling that causes clinically significant distress or impairment. Clinicians use the criteria below as part of a wider assessment, not as a checklist to label someone from a short bad period.
The criteria are about behaviour and consequences. They are not based only on how often someone gambles, the type of game used or whether a person has won or lost recently.
- Needing bigger bets gambling with larger amounts to get the level of excitement the person is seeking.
- Restless when cutting down becoming irritable or unsettled when trying to reduce or stop gambling.
- Repeated failed attempts to stop trying more than once to control, cut back or quit, without being able to maintain the change.
- Frequent gambling thoughts spending a lot of time thinking about past gambling, future bets or ways to get gambling money.
- Gambling during distress gambling while feeling distressed, rather than dealing with the underlying problem directly.
- Chasing losses returning to gamble after losing, in an attempt to recover what was lost.
- Hiding the extent of gambling lying to other people about how much gambling is happening.
- Risking major parts of life putting relationships, work, education or other important opportunities at risk because of gambling.
- Relying on others for money problems turning to other people to deal with serious financial consequences linked to gambling.
The American Psychiatric Association states that DSM-5 diagnosis requires at least four criteria during the past year, and its DSM-5 change summary says gambling disorder was moved into the substance-related and addictive disorders chapter. That placement reflects the manual’s classification.
It does not mean every person with gambling harm has the same pattern as every person with a substance use disorder. It also does not mean a screening score automatically becomes a diagnosis. The clinician still has to judge whether the pattern is persistent, problematic and linked to real impairment or distress.
Severity levels and exclusions
The American Psychiatric Association also uses severity specifiers and an exclusion rule in DSM-5. These guide clinical judgement after the criteria have been assessed. They help the clinician describe the current presentation more clearly, but they do not replace the full assessment.
Mild, moderate and severe
DSM-5 severity is based on how many criteria are met. Mild gambling disorder is linked to 4–5 criteria, moderate to 6–7 criteria and severe to 8–9 criteria. These levels do not measure a person’s worth, honesty or willpower. They describe the breadth of symptoms reported and observed during assessment.
A person with fewer criteria can still have serious financial, relationship or emotional harm that needs support. A person with more criteria may need a more intensive plan, especially if money pressure, mood symptoms or safety concerns are present.
When another condition explains it better
DSM-5 excludes gambling disorder when the gambling behaviour is better explained by a manic episode. That matters because mania can change risk-taking, sleep, spending and judgement across several areas of life, not only gambling. A clinician therefore asks about mood, energy, sleep, impulsivity and timing.
The aim is to identify the right clinical explanation and the right care plan, not to excuse harm or blame the person. If the gambling pattern continues outside the mood episode, the clinician may need to assess gambling disorder separately.
How ICD-11 describes gambling disorder
The World Health Organization’s ICD-11 lists gambling disorder under code 6C50. Its description focuses on impaired control over gambling, gambling taking priority over other activities, and continued gambling despite harm.
ICD-11 is a diagnostic classification used by health systems, so its wording is designed for coding and clinical records as well as assessment. The description is broad enough to cover different gambling formats, but the central clinical question is the same: has control reduced, and has gambling continued despite harmful consequences?
ICD-11 describes the pattern as typically evident over a 12-month period. It also allows a gambling disorder to be specified as predominantly online or predominantly offline. That distinction can matter in assessment because access, triggers and records may differ between app-based gambling, websites, retail venues and informal betting.
Online gambling can leave account histories, payment records and device-based triggers. Offline gambling may involve venue routines, cash withdrawals and travel patterns. The specifier helps describe the main route of gambling. It does not change the need for a full clinical assessment of control, priority and harm.
Problem Gambling Severity Index (PGSI)
The Gambling Commission describes the PGSI as a nine-item tool, with each item scored from 0 to 3 and a total score from 0 to 27. Ferris and Wynne developed it in the Canadian Problem Gambling Index, published in 2001 for the Canadian Centre on Substance Abuse.
The questions focus on harm and behaviour, such as betting more than intended, chasing losses, feeling criticised, borrowing or selling to gamble, and feeling that gambling may be a problem. The New York Council on Problem Gambling gives the following score bands.
-
0
Non-problem gambling
-
1–2
Low-level problems / low risk
-
3–7
Moderate-level problems / moderate risk
-
8
or more
Problem gambling
PGSI is often used in surveys, public health material and self-check tools because it is short and harm-focused. It can show that a conversation would be sensible, especially where gambling is already causing money pressure, secrecy, conflict or distress.
It still cannot rule in or rule out DSM-5 or ICD-11 gambling disorder. The NHS publishes these questions on its gambling help page. If you want to see how a questionnaire works without treating it as a diagnosis, use our gambling self-assessment test.
South Oaks Gambling Screen (SOGS)
A peer-reviewed study on PubMed Central describes the South Oaks Gambling Screen as a 20-item screen developed by Lesieur and Blume in 1987. A score of 5 or more is described as probable pathological gambler status.
That wording comes from older screening terminology and should not be read as a modern clinical diagnosis by itself. SOGS was built for screening, so a positive result means further assessment is warranted. It does not settle whether DSM-5 or ICD-11 criteria are met.
SOGS has been influential, especially in research, but it has a known limitation. The same peer-reviewed literature discusses concern that it can produce false positives in general-population surveys.
In plain terms, it may identify some people as likely cases who would not meet diagnostic criteria after a full clinical interview. This can happen because screening tools intentionally cast a wider net than a diagnostic interview.
That can be useful for finding people who need support, but it can also overstate disorder if the score is treated as final. SOGS is best read as a prompt for a fuller conversation, not a label.
Shorter screening tools
Some settings need a very short initial check. Short tools can help a clinician, helpline worker or support service decide whether a fuller gambling assessment is needed. Their strength is speed.
Their weakness is the same: fewer questions give less context. A short screen may miss the timing of gambling harms, the role of mood, whether gambling is mainly online or offline, and whether money pressure is creating immediate risk.
Tool | Questions | Typical use |
|---|---|---|
Lie/Bet questionnaire | 2 | A very short first check (Psychological Reports) |
BBGS | 3 | A short screen for gambling-related risk (Division on Addiction) |
NODS | – | NORC at the University of Chicago reports that 5 or more on the NODS indicates Type E / probable pathological gambler status under DSM-IV criteria. |
Why different tools can give different results
Two screens can point in different directions without either being useless. PGSI measures a recent pattern of gambling harms and risk. SOGS comes from a different research tradition and uses an older threshold. Lie/Bet and BBGS are designed to be short, so they trade detail for speed.
NODS is tied to DSM-IV criteria rather than DSM-5. A clinician should therefore ask what the tool was built to do, what time frame it covers and whether the result fits the person’s wider history.
This is the main reason screening scores should be handled carefully. A high score can make support more urgent, but the next step is still assessment. A low score may be reassuring, but it does not mean gambling is harmless if there are unpaid bills, relationship conflict, secrecy or distress.
A tool can only score the answers given. It cannot check bank records, speak to family members, identify a manic episode, review substance use or judge immediate safety. Those tasks belong in a clinical assessment.
What happens at a clinical assessment
A clinical assessment is usually a structured conversation. The clinician may use a screening score, but they should also ask follow-up questions and check whether the pattern meets DSM-5 or ICD-11 criteria. The focus is not only gambling frequency. Harm, control, urges, debt pressure, mood and safety all matter.
The assessment may also look at how gambling fits into daily routines: paydays, late-night use, sports events, venue visits, boredom, stress and access to credit. If gambling has affected limits, payments or time online, practical controls such as deposit limits and loss limits may be discussed alongside treatment and support.
- Gambling history what forms of gambling are involved, how the pattern changed and what tends to trigger it.
- Money and debts current pressure, borrowing, unpaid bills, concealed spending and any urgent financial risks.
- Relationships and work effects on partners, family, friends, employment, study and daily responsibilities.
- Mood, anxiety and other conditions symptoms that may have started before gambling problems or worsened alongside them.
- Alcohol or drug use substances that may increase risk-taking, reduce control or complicate treatment.
- Safety and self-harm thoughts any immediate danger, thoughts of self-harm or need for urgent support.
Preparation can make the appointment clearer. Useful material may include bank statements, a rough estimate of gambling losses, records from gambling accounts and a timeline of major changes. The timeline does not need to be perfect.
It only needs to help the clinician understand when gambling escalated, what was happening around that time and what has already been tried. Some people also bring notes about urges, attempts to stop, blocks used, relapses, debts, arguments or missed responsibilities.
If someone close to you is affected, our guide to helping a family member or friend with a gambling problem explains how to support without taking over.
A good assessment should feel specific, not moralising. The clinician is trying to separate gambling disorder from occasional risky gambling, financial stress, substance-related disinhibition, mood episodes and other mental health problems. They may ask direct questions because vague answers can hide risk. Direct questions are not accusations. They help build a plan that fits the pattern of harm.
After an assessment
An assessment should lead to a plan, not just a label. That plan may include psychological treatment, practical blocks, debt support, medication review for other conditions or referral to a specialist service. The right mix depends on the assessment findings and local care options.
It may also include reducing access to gambling accounts, asking someone trusted to help with financial safeguards, using self-exclusion, or arranging follow-up support. None of these steps guarantees recovery on its own, but each can reduce exposure to triggers while treatment addresses behaviour and underlying pressures.
For next steps, read how to stop gambling and what treatment options can involve. Family members can also seek guidance, even if the person gambling is not ready to attend support.
Where to get help
If gambling is causing harm, support options vary by country and service. Use our gambling helplines and support services by country page to find a relevant contact in your location.
Frequently asked questions
Is gambling addiction a recognised mental disorder?
Yes. The American Psychiatric Association recognises gambling disorder in DSM-5, and the World Health Organization lists gambling disorder in ICD-11 under code 6C50. Clinical services may use gambling addiction informally, but gambling disorder is the diagnostic term used in these manuals.
Can I diagnose myself with an online test?
No. An online test can show whether answers fall into a risk band, but it cannot make a clinical diagnosis. PGSI, SOGS, Lie/Bet, BBGS and NODS are screening tools, so a qualified clinician still needs to assess symptoms, context and exclusions.
What PGSI score indicates a gambling problem?
The New York Council on Problem Gambling gives a PGSI score of 8 or more as the problem gambling category. That result is a screening category, not a DSM-5 or ICD-11 diagnosis.
How long does a gambling disorder assessment take?
The length varies by service, clinical setting and complexity. A short triage call may be followed by a fuller appointment if risk, debt, mental health or safety concerns need more detail. Ask the service what to expect before the appointment.
Who can diagnose gambling disorder?
Diagnosis is usually made by a qualified mental health or addiction professional, such as a psychiatrist, clinical psychologist or addiction specialist. Local rules and service structures decide exactly who can record the diagnosis.