Pore-forming proteins (PFPs) possess the intriguing property that they can exist either in a stable water-soluble state or as an integral membrane pore. These molecules can undergo large conformational changes in converting between these two states. Much of what we know about how these proteins change their shape comes from work on bacterial toxins and increasingly, in more recent years, on toxins from other organisms. Surprisingly, a number of pore-forming proteins have recently been characterised that appear to have adopted similar stratagies to toxins for binding and inserting into biological membranes.