Cannabinoid receptor 1 (CB1R) antagonists have been proven to be effective anti-obesity drugs; however, psychiatric side effects have halted their pharmaceutical development worldwide. Despite the emergence of next generation CB1R blockers, a preclinical head to head comparison of the anti-obesity and psychiatric side effect profiles of the key compounds has not been performed. Here, we compared classical CB1R antagonists (rimonabant, taranabant, otenabant, ibipinabant, and surinabant) and non-traditional CB1R blockers (the partial agonist O-1269, the neutral antagonists VCHSR and LH-21 and the peripherally acting inverse agonist JD-5037) using an in vivo screening cascade. First, the potencies of these compounds to reduce CB1R agonist-induced hypothermia and decrease fasting-induced food intake were determined. Then, equipotent doses of the non-toxic compounds were compared in a diet-induced obesity (DIO) test, which includes measurements of metabolic syndrome markers. Psychiatric side effects were assessed by measuring anxiogenicity in an ultrasonic vocalization test. All classical CB1R blockers were centrally acting appetite suppressants and decreased body weight and food intake in an obesity-dependent manner, with only slight effects on metabolic syndrome markers. In addition, all classical CB1R blockers increased ultrasonic vocalization. Surprisingly, none of the non-classical CB1R blockers was eligible for the DIO comparison and side effect profiling. O-1269 and LH-21 induced convulsive behavior, whereas VCHSR and JD-5037 were devoid of any in vivo activity. The classical CB1R blockers displayed similar therapeutic and side effect profiles in vivo, whereas the available non-traditional CB1R blockers were not appropriate tools for testing the therapeutic potential of alternative CB1R inhibitors.