The concentration of grass‐pollen allergens in settled dust was measured over a period of 11 weeks in ten homes, five situated in an urban area and five in a rural area, in southwestern Finland during the summer of 1995. Dust samples were collected once a week, using a vacuum cleaner equipped with a special collection device (ALK, Copenhagen) and with a replaceable glass‐fiber filter in a filter dish. All dust samples were analyzed by means of an IgG‐ELISA‐procedure using murine monoclonal antibodies. The allergenic activity of grass pollen in settled indoor dust was low compared with that of birch pollen previously analyzed in the same area. Counts of airborne pollen used in comparison were obtained from the Burkard sampling station in Turku. The grass pollen season is not continuous in this area, but has sporadic relatively low peaks, because a variety of grass species flower at different times. Airborne grass pollen counts in Turku exceeded the threshold value of abundant (> 30 grains per cubic metre of air, Rantio‐Lehtimaki et al. 1991) only six times during the investigation period. The highest concentration of indoor allergenic activity was found about 3 weeks after the grass pollen peak. The allergenic activity of settled outdoor dust intercorrelated with the activity in settled indoor dust. This study supports the assumption that grass pollen antigens are carried home by people and pets. The concentration of grass‐pollen antigens resuspended in the air from settled dust may sporadically reach a level which provokes symptoms in allergic persons.