Background In addition to increasing the quality of life among concerned significant others (CSOs), Community Reinforcement and Family training (CRAFT) aim at helping CSOs motivate treatment-refusing identified patients (IPs) into treatment through a positive reinforcement process. The aim of the present study was to investigate if the following factors have an influence on IP treatment engagement1) Type of relation between CSO and the IP 2) The amount of time the CSO spend with the IP 3) if the IP knows that the CSOs seeks help, and 4) The CSO’s own alcohol use. Methods A secondary analysis from the Danish CRAFT study. CSOs completed a self-administered questionnaire at baseline, after three months, and six months. To investigate the relationship between the four variables and treatment engagement, logistic regression was used. Results CSO’s relation to the IP, the frequency of contact between the CSO and the IP, and the CSO’s AUDIT score were not associated with the IP’s treatment engagement. If the IP was informed that the CSO participated in CRAFT, the odds for IP treatment engagement was significantly higher (adjusted OR [(CI)] = 2.29 [1.13; 4.63] (p < 0.05), relative to the if IP not being informed. Conclusions CRAFT has a higher impact on the likelihood for treatment seeking, if the CSOs inform the IP about his or her own help seeking in order to change the situation. The underlying mechanism behind this is needs further investigations.