To be fair, he makes it clear in more recent books that he thinks you should invest most of your money conservatively and only use his get rich tricks on a small percentage of your portfolio. That said, most of his books describe agressive strategies that can make you (or lose you) a lot of money in the stock market in a short period of time. He particularly hypes two techniques: buying options on stocks that are about to split, and buying low priced stocks and selling options on them. His books are an entertaining read and make clear that it's possible to beat the market by a sizeable margin with the strategies he lays out. He's a little light on making clear the high amount of risk involved, so he's been taking a lot of heat on the net recently by people who lost their whole bankrolls on his strategies.
Graham's most famous book was one he co-wrote with David Dodd called Security Analysis. It's much deeper in it's mathematics, and seems to be intended as a textbook on analyzing stocks, as opposed to The Intelligent Investor, which is written for the general public.
This book has an interesting historical background on the way people have invested in past, with a particular focus on the mistakes they have made. He sort divides investors into those who invest based on firm foundations and those who build castles in the air. The former, of course, refers to the value type of investor, who bases their investments on things like P/E ratios. The latter refers to people who look for investments with a lot of hype, which generally are overpriced and run up to insane levels before collapsing (the idea being to get in when it's just starting to ramp up and get out before it crashes).
After describing both of these investment strategies, and historical background of their application, he goes on to assert that neither of them works worth a damn. Unfortunately, despite his prestigious position (he's an economics professor at Princeton), most of his refutations aren't well supported in the book. He mentions scientific studies to support his claims, but he doesn't cite most of the studies or give details of their setup, controls, etc. Granted, the book is for the general public, and they're not going to want to get into heavy statistics or economic arguments, but he could have at least cited the exact studies so the skeptical could check them out.
Thus, his system is to look for stocks with good earnings, good relative strength, a low market capitalization, a little industrial sponsorship (but not too much), and something new going for it, but only during bull markets. The newspaper really accentuates his system by putting two figures in their daily stock tables that no other paper has: earnings per share rank and relative strength rank. The two things that take any effort to find in his system are the 'N' and the 'M'. It takes a bit to find out about new developments in a company before the price takes it into account, but not all stocks skyrocket immediately on good news. Also, despite his encouraging words, we all know that noone can succeed at timing the market consistently. I think this is one of the best books for someone who is looking for a system for growth investing.
In How to Retire Rich, O'Shaughnessy distills the best information from that book to create a book that's written for the general public (as opposed to academics). Along with the standard rambling about why John Q. Public should invest in stocks instead of a savings account, he presents some of the best strategies from the previous book, and slight variations (including the popular Dogs of the Dow strategy). He also talks about whether to invest in bonds and so forth, but the meat of this book is in the strategies, the top of which is the aforementioned price to sales strategy under the guise of "Reasonable Runaways".