GREP: finding a string close to another

Here is an excerpt from a paragraph in my ID document:

xx £pe.$ u. £pa.$ yy

Using GREP, I should find and delete the first £ and $ signs, and then the second ones. The result should look like this:

xx pe. u. pa. yy

So I try to

Find

£(.+?)\$

Change

[nothing]

but this is only applied to the first occurence above, not to the second. I have no clue why this is so. Has anybody got an idea?

ID CS5 ME