Being able to measure and manage the carbon emissions of long-term digital preservation is an important aspect of environmental sustainability. Digital preservation is an active process involving the ongoing use of computer systems and data storage. This involves physical infrastructure and ICT equipment, the need for power and cooling, and a cycle of migrations and renewal. All of which generate carbon emissions, not just from energy consumption but also as a result of the embodied footprint from the manufacture, transportation and eventual disposal of the ICT equipment involved. But how big is this footprint? What aspects of digital preservation contribute most such as ingest or long-term storage? How do emissions depend on the type of content being preserved, what preservation actions are performed, and where preservation takes place such as in the cloud? Without being able to measure and calculate carbon emissions, it is hard to work out how best to reduce these emissions, or indeed whether other organisational activities offer lower hanging fruit. This poster explores these areas, presents a methodology for measuring carbon emissions of digital preservation, and gives some real numbers from large scale digital preservation in the cloud using real data for a range of digital preservation scenarios. The poster includes debunking a few misconceptions about the carbon footprint of energy consumption, especially in the cloud, it directs attention to the elephant in the room that is the embodied footprint of ICT equipment (servers and storage) that form the mainstay of digital preservation systems and infrastructure, and it provides some practical suggestions on how to reduce emissions. The overall aim of the poster is to help people understand how they can measure the carbon footprint of their digital preservation activities, how they can understand what aspects and activities contribute the most, and how they can make their own informed decisions on how to make effective reductions.