Objective: To model the potential financial implications of Australian programs supporting cognitively impaired community-dwelling older people. Methods: Markov cohort models of (a) an observational study of a residential dyadic training program for carers and people with dementia (GTSAH) and (b) a frailty intervention (FIT) in a cognitively impaired subgroup. Direct health and social welfare costs accrued over 5 years (2018 $AUD prices) were captured. GTSAH costs $3755, FIT costs $1834, and permanent residential aged care (P-RAC) costs $237 per day. Results: Modelling predicted costs break even in approximately 5 months for GTSAH and 7 months for FIT, after which these interventions saved funds. The primary driver of savings was the P-RAC cost (discounted at 5%/annum), at $121 030 for GTSAH vs $231 193 for standard care; and $47 857 with FIT vs $111 359 for standard care. Conclusions: Programs supporting cognitively impaired community-dwelling older people could be financially beneficial; further evaluation and implementation would be a worthwhile investment.