I'm finishing up development on a tourist information website that lists hotels, restaurants, bars, among other things for an up-and-coming tourist area in Cambodia: http://www.gosihanoukville.com/dining/index.html
Sections of the site that have listings are listed in a default order, and the visitor can choose to sort these listings on demand. There is also a section on the left of these pages that allow "filtering" of the results on the page. For a better example, on the restaurant page, there are filters for type of cuisine, location and budget. The cuisine filters at the top are the only ones that I want indexed - in case people search for "indian food in sihanoukville". The location and budget filters are just there for convenience, and could have been displayed with AJAX, but I wanted people to be able to bookmark them.
What I've done is set the meta robots tags on the pages I don't want indexed as "noindex, follow", and "index, follow" on all other pages. I have, however seen people use rel="nofollow" in their anchor tags. Which is the better practice? Or is it better to have them indexed anyway despite having duplicate content?