Since the issue of the metacognitive aspect of self-forgiveness has not been covered by previous research, this study examined the relationships of trait selfforgiveness, feeling of shame related to drinking, modified metacognitive model of depression and alcohol craving in a sample (N = 125) of inpatients during their treatment for alcohol dependence in psychiatric hospital. An extended metacognitive model of depression has been proposed to explain the association of these variables. The original metacognitive model was modified in such a way that depressive ruminations were replaced by ruminations related to drinking alcohol. The variables of trait self-forgiveness and feeling of shame related to drinking were added to model as antecedents, and finally alcohol craving as the outcome variable. Multiple mediation analyses with parallel and serial mediators were used to test the extended metacognitive model of depression previously disassembled into several smaller models. The model is largely validated. Self-forgiveness has an indirect effect on positive metacognitive beliefs about ruminations through feelings of shame related to drinking, and an indirect effect on ruminations about drinking through feelings of shame related to drinking and positive metacognitive beliefs about rumination. The direct effect of self-forgiveness on the interpersonal and social consequences of rumination is also significant. The effect of self-forgiveness on depression is fully explained by the mediator of rumination about drinking, and analyses also show an indirect effect through negative metacognitive beliefs about rumination. The indirect effect of self-forgiveness on alcohol craving through feelings of shame related to drinking and rumination about drinking was also discovered. All the above effects of self-forgiveness on the mentioned criterion variables are of negative sign. Ruminations about drinking have been shown to be the most significant predictor of depression and alcohol cravings. A modified metacognitive model of depression was also confirmed in a sample of individuals on treatment for alcohol dependence. The results suggest that forgiving oneself is a much more adaptive option for dealing with feelings of shame and avoiding all other potential negative consequences such as depression and alcohol craving than it is to repeatedly think about a perceived offense that has the opposite effect. In the second part of the study, the variables of changes in craving during treatment and trait self-forgiveness were examined as predictors of the number of days of abstinence 90 days after discharge from treatment. The repeatedly measured predictor of craving and trait self-forgiveness did not prove statistically significant in predicting the number of days of abstinence. This research provided new insights in understanding the constructs of selfforgiveness and feelings of shame by revealing their metacognitive context. It also contributes to a better understanding of depression in alcohol addicts by suggesting that the treatment of depression in these individuals should take into account the negative affect and rumination associated with the addiction itself. Also, the results suggest that for successful abstinence, it is not enough to be inclined to forgive oneself for one’s own transgressions, nor not to feel alcohol craving during treatment. When interpreting the results, the limitations of the research described in detail in the dissertation should certainly be taken into account.