The aim of this chapter is to describe a theme that bridges various topics within risk studies: the recursiveness of risk. There are several basic ways in which, when we describe some basic aspect of risk, that description becomes self-referential. A simple example of this, dealt with at more length later, is the idea that risks are transformed not solved by attending to them. The product of attending to a risk is another risk, and the product of attending to the other risk is another risk, and so on. A basic principle of the risk society ( Beck, 1992 ) is that modernity has to deal with the by-products of its own prior attempts to secure itself. So if you want to assess risk at any point you should assess the risk, as you see it, and then add in the risk of the response that the risk produces in you, and so on. You have to keep doing this until you have a reason to give up and call a halt to the analysis. The hope is usually that the risks caused by managing risks are smaller than the original risks. The reason for calling the process ‘recursive’ is that the idea that starts this chain (the potential for a harm) is the idea that we keep coming back to as we go down the chain. It is not just a long chain of connected possibilities, but rather a chain that keeps coming back to the same type of unit: a possible harm followed by a possible remedy followed by another possible harm. The importance of seeing the process as recursive is to be sceptical because it will ever be ‘solved’ to our full satisfaction. Once you believe that it is in the nature of risks to be things that produce more risks when they are attended to, by default you will have to treat risk management as an unending and evolving process over time, not as a two-step solution exercise.