Abstract
Using the scientific investigation of the Shroud of Turin as an extended example, it is argued that miracles are best understood not as violations of natural law, but as scientifically inexplicable events. It is then argued that even though we can imagine circumstances in which science itself might provide us with good grounds for believing that an event is scientifically inexplicable, these grounds would at best provide us with circumstantial evidence that the event was miraculous, and would in any case be inconclusive.