MEd is an annotation tool in which linearly-structured annotations of text or audio data can be created and edited. The tool supports multiple stacked layers of annotations that can be interconnected by links. MEd can also be used for other purposes, such as word-to-word alignment of parallel corpora.