Abstract Background Few studies have examined relationships between physical activity (PA) during mid-age and health costs in women. The aim of this study was to investigate associations between PA levels and trajectories over 12 years with medical and pharmaceutical costs in mid-age Australian women. Methods Data from 6953 participants in the Australian Longitudinal Study on Women’s Health (born in 1946–1951) were analysed in 2019. PA was self-reported in 2001 (50-55y), 2007 (56-61y) and 2013 (62-67y). PA data were linked with 2013–2015 data from the Medicare (MBS) and Pharmaceutical (PBS) Benefits Schemes. Quantile regression was used to examine associations between PA patterns [always active, increasers, decreasers, fluctuaters or always inactive (reference)] with these medical and pharmaceutical costs. Results Among women who were consistently inactive (< 500 MET.minutes/week) in 2001, 2007 and 2013, median MBS and PBS costs (2013 to 2015) were AUD4261 and AUD1850, respectively. Those costs were AUD1728 (95%CI: 443–3013) and AUD578 (95%CI: 426–729) lower among women who were consistently active in 2001, 2007 and 2013 than among those who were always inactive. PBS costs were also lower in women who were active at only one survey (AUD205; 95%CI: 49–360), and in those whose PA increased between 2001 and 2013 (AUD388; 95%CI: 232–545). Conclusion Maintaining ‘active’ PA status was associated with 40% lower MBS and 30% lower PBS costs over three years in Australian women. Helping women to remain active in mid-life could result in considerable savings for both women and the Australian government.