The PGSI: Scoring, Risk Categories, and Clinical Use

The Problem Gambling Severity Index (PGSI) is the scored core of the Canadian Problem Gambling Index, a longer survey that Jackie Ferris and Harold Wynne developed for the Canadian Centre on Substance Abuse and published in 2001. The full index asks about gambling habits, correlates and context. The nine PGSI items are the part that produces a number and a risk category.

It was built for population surveys, where it has been used widely since, and that origin shapes it. The questions were written for people in the general public who might never think of themselves as having a gambling problem, so they ask about concrete consequences: money, guilt, criticism from others, stress. That makes the PGSI easy to use outside a specialist clinic, in a first session or a general intake.

In the original report, internal consistency was good (Cronbach's alpha 0.84) and test-retest reliability was 0.78 (Ferris & Wynne, 2001). You can take a free PGSI screening here; it scores the nine items and returns the risk category.

Each of the nine items asks about the past 12 months and offers four answers: never (0), sometimes (1), most of the time (2) and almost always (3). Totals run from 0 to 27.

The items cover two kinds of ground. Some describe gambling behavior: betting more than the person could afford to lose, needing to bet larger amounts for the same excitement, going back another day to win back losses, and borrowing money or selling something to gamble. The rest describe consequences: feeling there might be a problem, being criticized for betting, health problems such as stress or anxiety, financial problems for the household, and guilt about gambling.

There are no reverse-scored items and no subscales in standard use. The total is the score.

0 (non-problem gambling): No consequences reported in the past year. This includes people who gamble now and then with no trouble, and people who don't gamble at all. A zero from someone in recovery from gambling is worth noting in the chart, because the score describes the past year and says nothing about history before it.

1 to 2 (low risk): A few consequences, usually endorsed at the sometimes level. Most people here have no gambling disorder. What matters is which items they answered: a 2 built from guilt and a single episode of overspending reads differently from a 2 built from chasing losses. A short conversation about how often they gamble and how much they spend is usually enough.

3 to 7 (moderate risk): Gambling is producing some harm. The person may be chasing losses some of the time, borrowing occasionally, or hearing concern from family. This is the band where brief intervention fits: naming the pattern, setting limits on time and money, and putting barriers in place before the score climbs. The gambling urges coping plan gives a client somewhere to write those barriers down.

8 or more (problem gambling): Gambling is causing clear harm, often with a loss of control. A score in this range calls for a full assessment against the DSM-5-TR criteria for gambling disorder, a look at debt and household finances, and screening for depression, substance use and suicide risk. Refer to a clinician who treats gambling disorder if that isn't your own area.

The PGSI is a screen. It sorts people by level of risk and does not diagnose gambling disorder. It also depends on what the person is willing to report, and gambling carries enough shame, and enough secrecy around money, that underreporting is a real risk. A low score next to a story of missing money or unexplained debt should be read with the story in mind.

The two middle categories are the weakest part of the scoring. Currie, Hodgins and Casey (2013) found strong evidence for the non-problem and problem gambling categories, but the low-risk and moderate-risk groups were hard to tell apart under the original cut points. They showed that a simple change to the scoring rules made those categories more valid. Summaries of their work usually give the revised bands as 1 to 4 for low risk and 5 to 7 for moderate risk, with 0 and 8 or more unchanged. This site's tool uses the original Ferris and Wynne bands.

The 12-month window also limits how often the PGSI is useful. Repeating it every few weeks measures overlapping years, so the score moves slowly. For tracking change during treatment, many clinicians repeat it every three to six months and use a weekly log of urges, bets and money spent in between.

DSM-5-TR defines gambling disorder as persistent, recurrent problem gambling that causes clinically significant impairment or distress, shown by four or more of nine criteria within a 12-month period. The criteria include needing to gamble with larger amounts to get the same excitement, restlessness or irritability when cutting down, repeated failed attempts to stop, preoccupation with gambling, gambling when distressed, chasing losses, lying to hide the extent of gambling, putting a relationship, job or opportunity at risk, and relying on others for money to get out of a financial hole. Severity is specified as mild (4 to 5 criteria), moderate (6 to 7) or severe (8 to 9).

Several PGSI items touch the same ground: betting larger amounts, chasing losses and borrowing to gamble all have clear counterparts among the criteria. Others have no direct match. The PGSI doesn't ask about withdrawal-like irritability, preoccupation or lying, and DSM-5-TR doesn't count guilt or health problems as criteria. The two also count differently. The PGSI sums how often each consequence happens, and DSM-5-TR counts how many criteria are present.

The practical result is that a PGSI score can't be converted into a diagnosis or a severity level. A score of 8 or more says a diagnostic interview is warranted. The interview is where gambling disorder is established, along with the rule-out DSM-5-TR requires: gambling that happens only during a manic episode is better explained by that episode.

Start with what the person answered, item by item. The items they endorsed most often point to where a plan should begin: chasing losses calls for a stop rule, borrowing calls for money barriers, criticism from family calls for a conversation about who already knows.

Money barriers carry a lot of weight in gambling recovery because the behavior runs on access to cash and credit. Self-exclusion programs at casinos and on betting sites, blocking software on phones and computers, and handing card control to a trusted person all add time between an urge and a bet. The gambling urges coping plan walks through these alongside urge surfing, delay, triggers, chasing and a plan for a slip, and the gambling disorder information sheet covers warning signs and treatment for a client who is new to the idea.

The National Problem Gambling Helpline from the National Council on Problem Gambling is free and confidential and answers by call, text or chat 24 hours a day: 1-800-MY-RESET (1-800-697-3738). It isn't a crisis line. If gambling losses or debt have brought thoughts of suicide, call or text 988.

What does a PGSI score of 5 mean? A 5 falls in the moderate-risk category under the original bands (3 to 7), so gambling is causing some harm and a brief intervention is a good fit. Under the revised bands from Currie and colleagues it also falls in moderate risk (5 to 7). Look at which items made up the five: a total built from chasing losses and borrowing is more urgent than one built from guilt alone.

Is a PGSI score of 8 a diagnosis of gambling disorder? No. An 8 is the start of the problem gambling category, and it says a full assessment is needed. Gambling disorder is diagnosed against the DSM-5-TR criteria in a clinical interview. Some people who score 8 or more meet those criteria and some don't.

Can someone who doesn't gamble take the PGSI? They can, and they'll score 0. In practice, many clinics ask first whether the person has gambled in the past 12 months, including lottery tickets, scratch cards, sports betting and online casino games, and give the PGSI to those who say yes.

Does the PGSI work for online and sports betting? Yes. The items ask about consequences of betting in general, with no reference to a particular game or venue, so they apply to phone-based sports betting and online casinos as well as to in-person gambling.

American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders (5th ed., text rev.). https://doi.org/10.1176/appi.books.9780890425787

Currie, S. R., Hodgins, D. C., & Casey, D. M. (2013). Validity of the Problem Gambling Severity Index interpretive categories. Journal of Gambling Studies, 29(2), 311-327. https://doi.org/10.1007/s10899-012-9300-6

Ferris, J., & Wynne, H. (2001). The Canadian Problem Gambling Index: Final report. Canadian Centre on Substance Abuse.

National Council on Problem Gambling. National Problem Gambling Helpline, 1-800-MY-RESET (1-800-697-3738). Free and confidential, by call, text or chat, 24 hours a day, 365 days a year.