Falls are the most common cause of injury in older people, with consequences for the individual and society. With an increasing population of older people, falls and related costs are expected to increase. It is crucial to identify scalable and cost-effective interventions and subsequently reduce fall-related costs. The aim was to evaluate the cost-effectiveness of the Safe Step digital fall preventive exercise intervention over a period of 12 years and, in addition, to evaluate the impact of increased recruitment cost and decreased intervention effect. The intervention was evaluated in an observational study in a municipality context targeting community-dwelling older people of age 70 +. A Markov model with five states was used to model the cost-effectiveness of the Safe Step intervention and evaluate quality-adjusted life years (QALYs) and fall-related costs from a societal perspective. By using data from a meta-analysis as basis for the estimated intervention effect, the Safe Step intervention was compared with a no-intervention alternative. The results showed that the Safe Step intervention dominated no intervention. In the sensitivity analysis with the most conservative estimate of intervention effect, the ICER was €7 616 per QALY gained. Hence, Safe Step showed to be a cost-saving fall preventive intervention in older people at risk of falling and potentially cost-effective even with a low estimated intervention effect. Future studies on efficacy of fall preventive digital interventions will contribute in precising effect estimates and enhance the validity of these cost-effectiveness results.