Growing interest in ecosystem restoration has recently turned the focus on tree planting, one of the most widely used restoration tools globally. Here, we study the restoration potential of tree planting in a cool‐temperate forest in Shiretoko National Park, northern Japan. We used simulation modeling to investigate the long‐term success of tree planting in restoring biodiversity and the climate change mitigation function relative to intact natural forests. Specifically, we investigated 31 different restoration scenarios, consisting of five planting densities (1,000–10,000 trees/ha) × six levels of planted tree species richness (one to six species) + one no‐planting scenario. We examined these scenarios at different distances from natural forests serving as a seed source (0–300 m) to quantify the potential for natural regeneration. In restoration areas in close proximity to a natural forest, species‐rich high‐density planting scenario performed best, reaching >50% of the reference values from intact natural forests within 33 years for both restoration goals. However, variation in restoration outcomes was small when >2,500 trees/ha of more than four species were planted, regardless of distance to seed source. In contrast, biodiversity restoration was considerably delayed in scenarios where planted species richness was low as well as in restoration areas that were far from a seed source yet relied solely on natural regeneration. We here demonstrate how forest landscape simulation can be used to identify viable restoration options for managers across multiple restoration goals as an important step to bridge the research–implementation gap in forest restoration.