Background and aims: The Prevention Paradox (PP) suggests that a large proportion of aggregate harm from gambling occurs to people who do not have a gambling disorder. However, it has not yet been tested using a population-representative sample. We aimed to test whether the PP applies to gambling in Finland. The prevalence rates of diverse harmful consequences from gambling were surveyed amongst a population-representative sample of past-year gamblers. Methods: The study used first wave data (N = 7,186) of Finnish Gambling Harms survey, collected via online and postal surveys in 2017. A subset of 3,795 adults (≥18 years), who had gambled at least monthly in 2016, were selected for analysis. Measurements: Gambling-related harms were evaluated with the 72-item Harms Checklist. Problem and Pathological Gambling Measure (PPGM) measured respondents' probable disordered gambling from the subset of items for impaired control (4 questions) and other issues (3 questions). Findings: Consistent with previous findings, the majority of harms were reported by those in the less severe PPGM categories (i.e. scoring <5). However, considering each domain separately, this was true only for financial, emotional/psychological, and work/study harms. The PP was not supported for health, relationship, or social deviance harms. Conclusions: The population prevalence of the most serious harms (e.g. unsafe living conditions) is concentrated among those with severe impaired control issues. However, even excluding the ∼15% of harms occurring to occasional gamblers, most financial, emotional and work/study impacts occur to those with lower levels of control issues. Efforts at harm reduction should focus on the entire spectrum of issues that people experience from their gambling.