Impulsivity is an important factor in adverse outcomes such as substance use, problem gambling and psychopathology. Extensive research has shown these negative outcomes are associated with both self-report and behavioural measures of impulsivity but these two measurement domains are not themselves associated. There has been limited research in prison samples. This is surprising given the high variability in impulsive behaviours that should make them ideal for investigating the convergence of impulsivity measures. Using a cross sectional design we investigated the associations of impulsivity – measured by self-report and two behavioural indices - with substance misuse and psychopathology in a sample of 72 male prisoners. We found higher self-reported impulsivity was associated with crack/cocaine use, problem gambling and a positive screen for personality disorder. Behavioural measures of impulsivity showed fewer associations with problematic behaviours; they were also not independent predictors of impulsive behaviour in multivariate analyses. These data suggest that self-reported impulsivity is a more consistent predictor of problematic behaviours than behavioural measures in a sample of people with significant levels of substance use and psychopathology. This difference could reflect relevance of self-reported measures to emotionally charged decision-making in daily life compared to more neutral behavioural measures.