Distortion in NLP is one of the three major processes (including Deletion and Generalisation) on which the Meta Model is based. Distortion occurs when something is mistaken for that which it is not, when things are incorrectly included in our Internal Representations.
For instance, when you are looking for a certain piece of wire in a cupboard and you thought it was a grey colour, the wire you are looking for may not be there. But when you just looked, you found it (because it actually wasn’t grey). Or you see a shake moving in a desert and you run because you think snake! But it was a piece of discarded rope.