The synthetic tetrahydrocannabinol-analog nabilone improved non-motor symptoms (NMS) in Parkinson's disease (PD) patients in a placebo-controlled, double-blind, parallel-group, randomized withdrawal trial with enriched enrollment (NMS-Nab-study). This was a single-center open-label extension study to assess the long-term safety and efficacy of nabilone for NMS in PD. To be eligible for this study, patients had to be treatment responders during the previous NMS-Nab-trial and complete its double-blind phase without experiencing a drug-related serious/severe/moderate adverse event (AE). Patients were re-introduced to nabilone during an up-titration phase until their overall NMS burden improved. Nabilone was continued for six months with clinic visits every 3 months. Evaluation of AEs was based on self-report and clinical assessment. Twenty-two patients participated in the NMS-Nab2-study (age-median 68.33 y, 52% females, disease duration-median 7.42 y). Nabilone was well tolerated with concentration difficulties as the most common treatment-related AE (possibly/not related n = 1 each). One in two drop-outs discontinued because of an AE for which a prohibited concomitant medication needed to be introduced (night-time sleep problems). Efficacy evaluation showed a significant and lasting improvement in NMS burden according to the CGI-I (79% at V3). Nabilone improved overall sleep (NMSS Domain-2: -8.26 points; 95%CI -13.82 to -2.71; p = 0.004; ES = -0.72), night-time sleep problems (MDS-UPDRS-1.7: -1.42 points; 95 CI -2.16 to -0.68; p = 0.002; ES = -0.92), and overall pain (KPPS Total Score: -8.00 points; 95%CI -15.05 to -0.95; p = 0.046; ES -0.55 and MDS-UPDRS-1.9: -0.74 points; 95%CI -1.21 to -0.26; p = 0.008; ES = -0.74). This study demonstrates continuous long-term safety and efficacy in PD patients responding early to nabilone without intolerable side effects.