Analysis of long‐term trends in forest carbon stocks is challenged by interactions among climate change, wildfire and other disturbances, forest management actions, and heterogeneous vegetation responses. For such circumstances where complex interactions make it difficult to encompass the full range of processes in any one mode of analysis, expert elicitation is a well‐developed method for documenting judgments about uncertainty, based on available evidence, to inform ongoing decision‐making. Applying this method for the Sierra Nevada, we evaluate subjective probabilistic estimates of trends in aboveground forest carbon for different management scenarios toward the goal of maximizing carbon stored, while also considering implications for wildfire risk. The analysis examines the effects of four treatments in isolation (thinning, timber harvesting, prescribed burning, managed wildfire), as well as a user‐defined management portfolio allocating resources across five management practices (thinning, harvesting, prescribed burning, firefighting, and restoration). The expert elicitation suggests that aboveground forest carbon stocks will decline 8%, from 126 to 116 tC/ha, between 2030 and 2100 (median estimate across experts) assuming conventional forest management practices are continued. Out of all surveyed practices, the custom user‐defined management portfolio results in the highest carbon stock of 129 tC/ha which is 11% higher than conventional practice in 2100 at the 50th percentile. The expert elicitation indicates less beneficial carbon sequestration outcomes than recent modeling studies. Suggesting co‐benefits across objectives, 75 experts collectively estimate a 61% likelihood that managing for carbon also reduces wildfire risk. By contrast, decreases in carbon stocks are anticipated for large magnitudes of climate change or substantial decreases in forest management investments.