Objectives Problem gamblers in treatment are known to be at high risk for suicidality, but few studies have examined if this is evident in community samples. Evidence is mixed on the extent to which an association between problem gambling and suicidality may be explained by psychiatric comorbidity. We tested whether they are associated after adjustment for co-occurring mental disorders and other factors. Study design Secondary analysis of the Adult Psychiatric Morbidity Survey 2007, a cross-sectional national probability sample survey of 7403 adults living in households in England. Methods Rates of suicidality in problem gamblers and the rest of the population were compared. A series of logistic regression models assessed the impact of adjustment on the relationship between problem gambling and suicidality. Results Past year suicidality was reported in 19.2% of problem gamblers, compared with 4.4% in the rest of the population. Their unadjusted odds ratios (OR) of suicidality were 5.3 times higher. Odds attenuated but remained significant when depression and anxiety disorders, substance dependences, attention-deficit/hyperactivity disorder, and other factors were accounted for (adjusted OR = 2.9, 95% confidence interval = 1. 1, 8.1 P = 0.023). Conclusions Problem gamblers are a high-risk group for suicidality. This should be recognised in individual suicide prevention plans and local and national suicide prevention strategies. While some of this relationship is explained by other factors, a significant and substantial association between problem gambling and suicidality remains.