Huntington's disease (HD) is a rare multifactorial neurodegenerative disease. Both its natural course and any placebo effect are poorly known. All are obstacles to design randomized controlled trials (RCTs). We conducted meta-analyses of RCTs and cohorts on all parameters of the Unified Huntington's Disease Rating Scale to determine the most appropriate outcomes and to minimize the number of patients required to design RCTs in HD. Twenty-four RCTs were included, involving 838 patients with a mean age of 50.0 ± 2.3 years and a mean total functional capacity (TFC) score of 9.8 ± 0.6. Nineteen cohorts were included involving 1939 patients with a mean age of 48.9 ± 2.3 years and a mean TFC of 10.1 ± 0.7. Significant deterioration was observed in RCTs for all scores except behavioral score. Effect sizes were comparable between RCTs and cohorts for each test except that there was a significant difference for TFC. The weighted mean deterioration per year on the TFC scale was −0.5 (0.2) in RCTs and −0.8 (0.2) in cohorts. The lowest number of patients required per group in a RCT was for TFC (19 per group), whereas 30 patients would be required per group for the total motor score (TMS). For cognition, the verbal fluency test required the smallest number of patients: 104 per group. In conclusion, TMS and TFC are the most appropriate outcomes to design RCTs on HD likewise the verbal fluency test for cognition. Our results suggest an effect of placebo administration on the total functional capacity.