A two-part sediment sampler (stationary base and removable trap) was designed for a long-term study of stream sedimentation associated with highway construction. Before the long-term study, a laboratory study in an experimental flume examined efficacies of our sampler and two other sediment samplers: a modified core sampler and Whitlock-Vibert boxes. Based on the flume experiment, the efficacy of our sediment sampler was consistent with that of core and Whitlock-Vibert samplers. The advantage of our two-part sediment sampler design is that it allows for repeated removal of sediment samples without continual disturbance of the streambed. It also minimizes labor necessary to collect sediment samples. Our sampler is designed for long term monitoring of streams impacted by sedimentation and not for characterization of stream substrate composition.