Personal respirable (N = 230) and total (N = 237) dust measurements were made in two coastal British Columbia sawmills using a sampling strategy that randomly selected workers from all jobs in the mills over two seasons. Information about job title, department, season, weather conditions, location of the job relative to wood-cutting machines, and control measures also was collected at the time of sampling. Only 16 respirable wood dust samples were above the detection limit of 0.08 mg/m3; all 16 had levels < or = 0.20 mg/m3. Total wood dust concentrations were also low (36% less than the detection limit), with a mean of 0.51 mg/m3, and ranging from < 0.08 to 52 mg/m3. Measurements of exposure taken close to chippers, planers, and multiple saws had the highest total wood dust levels. Sawmill department and booth enclosures also were associated with wood dust concentrations, while local exhaust ventilation and weather conditions were not. Wood dust levels in this study were generally lower than in other studies of this industry, but most sawmill investigations report mean wood dust concentrations lower than those measured in the furniture and cabinetmaking industries, where concerns about wood dust exposures initially were raised.