A number of studies have shown that stress and an activated hypothalamic-pituitary-adrenal (HPA) axis are associated with increased voluntary alcohol drinking. Recently, associations have been found between activated HPA and hypothalamic-pituitary-gonadal (HPG) axes in alcohol-preferring AA and non-preferring ANA, F2 (crossbred second generation from original AA and ANA), and Wistar rats. The aim of the present study has been to determine the role of corticosterone and alcohol-related testosterone-effects in subsequent alcohol drinking in AA, ANA, F2 and Wistar rats. The present study comprises of four substudies presenting new analyses of existing data, by which correlations between basal corticosterone levels, changes in testosterone levels during alcohol intoxications and subsequent voluntary alcohol consumption are investigated. The results displayed positive correlations between basal corticosterone levels and subsequent alcohol-mediated testosterone elevations, which was positively associated with voluntary alcohol consumption. The results also showed a negative correlation between basal corticosterone levels and alcohol-mediated testosterone decreases, which was negatively associated with alcohol consumption. In conclusion, the present study displays novel results, according to which the HPA axis, one hand, relates to testosterone elevation (potentially causing and/or strengthening reinforcement) during alcohol intoxication, which in turn may relate to higher voluntary alcohol consumption (AA rats). Vice versa, the HPA axis may also relate to alcohol-mediated testosterone decrease (causing testosterone reduction and disinforcement) and low-alcohol drinking (ANA, F2 and Wistar rats). In addition, the present results showed that alcohol-mediated testosterone changes may also, independently of the HPA axis, correlate with voluntary alcohol drinking, which indicate the impact of genetic factors. Thus, the role of the HPA-axis may be more related to situational stress than to intrinsic factors. In further studies, it should be investigated, whether the present results also apply to stress and human alcohol drinking.