To evaluate and compare the efficacy and safety of different doses of inhaled levodopa (CVT-301) for the management of OFF episodes in patients with Parkinson’s disease using a network meta-analysis.

CVT-301 provides rapid relief of OFF episodes, but the relative efficacy of different doses remains uncertain. Network meta-analysis (NMA) allows dose-level comparisons that are clinically informative.

We systematically searched PubMed, EMBASE, SCOPUS, and Web of Science till September 2025 for randomized controlled trials evaluating CVT-301 in Parkinson’s disease. Outcomes included achievement of ON state, UPDRS-III score, Patient Global Impression of Change (PGIC), and adverse events. Data were pooled using random-effects network meta-analysis (netmeta, R), assessing consistency and heterogeneity across networks. Language refinement was assisted by ChatGPT (OpenAI, GPT-5).

Five randomized controlled trials (RCTs) were included. For the ON state (k = 4, n = 625), CVT-301 35/50 mg three times daily with dose escalation was the most effective regimen (OR 6.41, 95% CI 2.27–18.09; p = 0.0004), followed by 60 mg (OR 2.28, 95% CI 1.28–4.06) and 84 mg administered up to five times per day (OR 2.15, 95% CI 1.20–3.84). UPDRS-III outcomes (k = 3) showed the greatest improvement with 35/50 mg (MD 9.00, 95% CI 8.56–9.44; p < 0.0001). PGIC analyses (k = 2, n = 271) similarly favored 35/50 mg and 84 mg up to five times daily. In terms of safety, there were no significant differences in the incidence of falls or nausea. However, respiratory infections occurred more frequently with the 84 mg dosing regimen (RR 3.33, 95% CI 1.22–9.10).

CVT-301 35/50 mg provides the most consistent efficacy, while higher doses such as 84 mg increase respiratory risk. Dose-specific analyses may better inform clinical use of inhaled levodopa in Parkinson’s disease.