Dementia is a global public health priority and risk reduction is an important pillar of the public health response. While 40% of cases are estimated to be attributable to modifiable health and lifestyle risk factors, public awareness of the evidence is low, limiting peoples’ opportunity to adopt risk-reducing behaviours. To address this gap, we designed, implemented, and evaluated an educational intervention, the Preventing Dementia Massive Open Online Course (PDMOOC). This mixed-methods study examined the reach and impact of the free and globally available PDMOOC, to assess its potential to provide effective dementia risk reduction education to a broad international audience. Over 100,000 individuals participated in the PDMOOC across seven iterations from 2016 to 2020, with 55,739 of these consenting to participate in research. Their mean age was 49 years (SD = 15), they came from 167 different countries, and the majority were female (86%), had completed post-secondary education (77%), lived in high-income countries (93%) and worked in health care and social assistance (63%). This demographic profile changed across time, with more men, people with higher education and people from low- and middle-income countries participating in recent course iterations. Two-thirds of participants completed the PDMOOC; completion was associated with being aged 50 to 70 years, residing in a high-income country, having tertiary education, and working in the health sector. Participants reported high levels of satisfaction with the PDMOOC, improved dementia risk reduction understanding and self-efficacy, increased motivation to maintain healthy lifestyles, and, importantly, application of their learning to health behaviour change with the potential to reduce their dementia risk. The PDMOOC educated a large global audience about dementia risk reduction, which contributed to participants making risk-reducing behaviour changes. This suggests MOOCs can be a successful public health strategy to improve dementia risk reduction understanding.