Several studies have shown that gambling problems are cyclical but few have empirically investigated factors that are associated with change. The purpose of this article is to prospectively examine associations between life events and problem gambling severity in a cohort of gamblers. Occurrence of life events and gambling problem severity were assessed 3 times over a period of 2 years in a cohort of nonproblem and problem gamblers (N = 179) drawn from a representative sample derived from a population survey. Cross-lagged analyses revealed that cumulative number of life events were associated with an increase in severity of problem gambling 12 months later. Regression analyses showed that significant life events in several domains, for example, "change in sleeping habits," "accidental injury or illness" or "retirement," are likely to be associated over time to the increase or the continuation of risky gambling habits. This study's findings on the potential negative influence of cumulative number of life events, or of specific ones, are informative for secondary prevention and treatment.