An age-structured projection model was used to study the impacts of alternative exploitation intensities on geoduck populations, based on a large accumulation of survey data, age composition data and published estimates of natural mortality. Data were analyzed and results presented by geographic region. Historic recruitment patterns were back calculated using an age-structured model. Trends in recruitment were independent of the value of M used, although rates were higher when M of 0.036 was applied compared with M of 0.016. Historic recruitment rates were found to be highly variable in the prefishery state: rates generally increased from the early 1930s to 1950, decreased until early 1960, increased to another peak in the mid 1960s, declined until the mid 1980s and have been increasing to present. Future recruitment was simulated from the estimated historic pattern. The fishing patterns examined were combinations of different mortality rates (0.016 or 0.036) and different periods of historical recruitment (beginning in 1940 or 1960). For each simulation year, in each set of 1,000 runs, the ratio of current biomass to virgin biomass was calculated and compared with the management objective of not exceeding 50% of virgin biomass within 50 y of harvest. An exploitation rate of 1.2% and 1.8% of estimated current biomass is recommended for the west coast of Vancouver Island and the rest of the coast respectively.