We examined the effect of different invitational policies on the reduction of breast cancer mortality at 60-79 years of age within the Finnish mammography programme in 1992-2003, which varied in its coverage at 60-69 years of age. The data from 260 municipalities were grouped into three categories: regular invitations at 50-59 years of age only, regular invitations at 50-69 years of age, and regular invitations at 50-59 years of age with irregular invitations at 60-69 years of age. Observed deaths from breast cancer were compared to those expected without screening among all women and among the screened and non-screened women. Observed deaths were obtained from population data and from a cohort follow-up in 1992-2003. Expected deaths were derived by modelling breast cancer mortality at population level in 1974-1985 and 1992-2003. The reduction in breast cancer mortality was strongest, 28% (0.72, 0.51-0.97), in municipalities with regular invitations at 50-69 years of age. No overall effect at 60-79 years of age was observed with regular invitations at 50-59 years of age. The study confirms a reduction by screening of breast cancer mortality in Finland. Uniform extension of invitations to 60-69 years of age would increase the number of prevented breast cancer deaths among the elderly.