Cannabis use is more common in individuals with non-affective psychotic disorder and their siblings compared to healthy controls. As cannabis use is associated with a greater risk to develop psychotic disorder and an adverse outcome in those who already developed psychosis, it is important to know the role of craving in continued cannabis use and relapse in these vulnerable subjects. Therefore, we examined the validity of the Obsessive Compulsive Drug Use Scale for cannabis (OCDUS-CAN) in patients with non-affective psychotic disorder, their siblings, and healthy controls who all used cannabis in the past year. Simultaneous component analysis (SCA) was used to determine component weights that optimally explained the (co)variance of the OCDUS-CAN variables in these different populations simultaneously. A three-component SCA solution explained 74.2 % of the total variance, and consisted of well-interpretable subscales that could be best described as craving/urge, resistance, and impact. Reliability of the subscales was good. The three subscales significantly discriminated between frequent and infrequent cannabis users. Patients scored higher on the craving/urge and impact scale than siblings and controls, which could be related to primary and secondary symptoms of their disorder. The OCDUS-CAN is well suitable for people with or without vulnerability for psychotic disorder.