The "snippet" that Google displays below a found page title is taken from DMOZ's Open Directory Project (ODP) or the META DESCRIPTION tag on the page, or from the page contents itself:

If your web site is listed in the ODP then Google (and MSN) will regularly use the ODP's description text to describe your web site in their search results. This can be problematic as ODP descriptions are often woefully out of date, not reflecting your current (read: preferred) messaging. It used to be the case that you were forced to suffer this anachronistic data being returned, and take it on the chin...

...that was until May of this year (2007).

Google did the decent thing and introduced a Google-specific META robots tag value. The NOODP (No Open Directory Project) robots tag prevents Google from using DMOZ's description. There is a comprehensive practical instruction from Google on how to implement it here, but as a brief overview here are the basics:

To prevent all search engines (that support the meta tag) from using this information for the page's description, use the following:

• <META NAME="ROBOTS" CONTENT="NOODP">

To specifically prevent Google from using this information for a page's description, use the following:

• <META NAME="GOOGLEBOT" CONTENT="NOODP">

If you use the robots meta tag for other directives, you can combine those. For instance:

• <META NAME="GOOGLEBOT" CONTENT="NOODP, NOFOLLOW">

Adding one of these meta tags your pages won't have an instant effect, you'll have to wait with baited breath to be spidered!

The added benefit is that by using this tagging method Google will not only ignore your ODP description, but the ODP title too.