We used radio telemetry to monitor passage and describe behavior characteristics of adult Pacific lampreys, Entosphenus tridentata, during their upstream migration at the Willamette Falls Project (Project) on the Willamette River near Portland, Oregon. Our objectives were to document: (1) specific routes of passage at the dam and falls; (2) duration of passage through different routes; and (3) overall passage success. During the spring through autumn of 2005 and 2006, fish were captured in a trap located in the fishway at the Project or collected by hand from the falls, surgically implanted with a radio tag, and released 2 kilometers downstream of the Project. We radio tagged 136 lampreys in 2005 and 107 in 2006. In both years, more than 90 percent of the fish returned to the Project with a median travel time of 7-9 hours. Most fish were first detected at the Project from about 20:00-23:00 hours. In 2005, 43 fish (35 percent) successfully passed through the fishway of the Project, which has four separate entrances and three distinct passage channels or legs that converge at one exit. Prior to the installation of flashboards around the perimeter of the falls in July, lampreys used all three legs of the fishway to pass the Project. After flashboards were installed, only fishway leg 1 was used. The peak of passage occurred in August. No fish passed over the falls, but 13 percent of the lampreys that traveled to the Project ascended at least partway up the falls. In 2006, 24 fish (23 percent) passed the Project, again primarily using fishway leg 1. Most fish passed prior to June 9 when the powerhouse was shut down due to construction. Although 19 lampreys ascended the falls, only 2 passed through this route in late June and early July. Flashboards were not installed in 2006. For both years, the time it took for fish to pass through the fishway depended on which leg they used - the median passage time was at least 4-5 hours in fishway legs 2 and 3 and ranged from 23 to 74 hours in fishway leg 1. Many fish resided in the tailrace for times ranging from a few hours to almost a year and eventually left the Project and moved downstream. Collectively, our results indicate that passage of radio-tagged upstream migrating Pacific lamprey at the Willamette Falls Project is relatively poor compared to passage success of these fish at dams on the Columbia River. Factors contributing to the low passage of lampreys at the Project may include low flows and water levels at fishway entrances, impediments in the fishways, delayed tagging effects, changing environmental and operational conditions, a learned aversion to a fishway, difficult passage over the falls, or not all lamprey are destined to migrate upstream of the falls.