For some reason I find myself subscribed to more than thirty magazines, and I'm not sure how that happened. I'm certainly not able to keep up. But I did want to mention which ones I find especially interesting in terms of discovering new books, in order of importance.
The New Yorker: this is a great magazine that appears weekly, with in depth articles, short fiction, and reviews of a number of books (not the mass market kind). I've found selections from various authors such as Sherman Alexie and David Sedaris, which led me to look for their books. They also feature great poetry excerpts. A favorite...
Smithsonian: The science journal isn't nearly as dry and dull as Discover. The articles are not dumbed down, and they are long enough to get a real feel for what the topic is, from Hadrian's Wall to Forensic Astronomy (fascinating stuff). The scope of articles is large and again, many great titles can be found in this monthly journal.
Kiwi: This is for moms, with earth friendly tips and links to realistic parenting (yes, you have permission to build a fort rather than do dishes. Sometimes.). A great selection of unusual toys, multicultural crafts, and best of all, reviews of children's books that are not the typical selections.
Oprah: She's a love/hate person for most people, but I've found some great titles in this magazine in her Reading Room section. Without this feature, I would have never discovered Cloudstreet, Revolutionary Road (before the movie buzz!), House of Meetings, etc. Not all her titles are mass market, and although I don't follow her book club titles (too depressing!), I do enjoy hearing comments on many of the other books.
Time: I read this mostly for Joel Stein's column, but it's a great classic general news source. The articles are a bit too short to be deep, but it makes for great general variety. The Economist is even better.
Then there's the "don't bother" category: New York Review of Books seems like overkill and leaves you slightly dizzy. The Strand focuses on mystery titles exclusively, and seems formulaic.