Climate policy uncertainty (CPU) has gained prominence in renewable energy research, yet its determinative role remains largely overlooked. This study intends to bridge this gap by exploring whether CPU in the USA affects renewable energy consumption (REN), with a focus on the period 2000–2021. Employing the “Bayesian model averaging approach,” it first identifies the key drivers of REN, accounting for competing theories and empirical evidence. The findings underline not only the importance of economic activity, population, and research and development as factors influencing REN, but also the role of CPU. The “autoregressive distributed lag bounds testing approach” is then used to assess their short- and long-run impacts on REN. The results show that CPU negatively impacts REN, and this impact is even more pronounced in the long run. Consequently, the transition to renewable energy as a climate solution faces challenges due to climate change sensitivity and unclear energy policy. The negative effect is also observed in gross domestic product, industrial production, and oil prices, reflecting the fact that the US economy requires a large, stable, and controllable amount of energy currently unmet by renewables. Encouragingly, intensified international trade and research and development activities foster competitiveness and technological progress in REN. The key implication is that policymakers must prioritize stable climate policies to stimulate sustainable investments in renewable energy. Graphical abstract: [Figure not available: see fulltext.].