Problem gambling is conventionally defined by the score in a specific questionnaire exceeding some critical value and data suggests is that 0.7% of adults in the UK could be afflicted. However, the literature has not evaluated the size of the harm associated with such an affliction and this research evaluates the effect of problem gambling on self-reported well-being which, together with a corresponding effect of income on well-being, allows us to construct a money-metric of the (self) harm associated with being a problem gambler. Our estimates suggest that problem gambling imposes a very large reduction in individual well-being.