To investigate the effectiveness of closed‐type check dams on retaining driftwood, we conducted a series of wood‐laden debris flow flume experiments. In the experiments, we varied (1) the initial sedimentation condition of the check dam and (2) the length of input driftwood, investigating the amount of driftwood retained by the check dam and its temporal change. As a result, we found that the retention mechanism of the driftwood is more strongly governed by the difference in the initial sedimentation condition rather than by the difference in driftwood length. The ratio of retained driftwood to supplied driftwood was generally higher under the empty‐dam condition, by which more than half of the driftwood was retained regardless of experiment cases. The ratio was notably higher under the empty‐dam condition, with driftwood length longer than the spillway width. Under the filled‐dam condition, on the other hand, the ratio showed relatively wider fluctuation among experiment cases, even when the driftwood length was longer than the spillway width. In light of the log behavior behind the check dam under all cases, driftwood was more certainly retained due to the resilient blockage structures formed by driftwood near the check dam spillway. In the case of the empty‐dam condition, driftwood at the head of the debris flow that reached the check dam was covered and immobilized by the following sediment. Further efforts should be made to evaluate other factors, e.g., congestion level of log transport that potentially affect the effectiveness of closed‐type check dams. Our findings suggest that maintaining the storage capacity of closed‐type check dams is essential for utilizing existing structures for mitigating driftwood disasters.