This dataset provides tabular data output from a series of modeling simulations for forest ecosystems of the U.S. state of California under different initial conditions scenarios. We used the LUCAS state and transition simulation model with carbon stocks and fluxes based on the Carbon Budget Model of the Canadian Forest Sector (CBM-CFS3) to simulate changes in forest ecosystem carbon balance resulting from historical land use and land cover change, annual climate variability, and disturbance from wildfire and drought-induced forest die-off. The model was run at a 1-km spatial resolution on an annual timestep for the years 1985 to 2020. We simulated 36 initial conditions scenarios based on unique combinations of spatial datasets used to define forest extent, forest composition and forest age at the beginning of the simulation period. Results presented here have been aggregated from the individual grid cell level and summarized for the entire state of California.