I seem to remember from a measure theory class a long time ago something about the following result. My memory may be totally wrong.

Is it true that on the real line with the ordinary lebesgue measure, any set of positive measure is the union of intervals?