Yes, there are lots of chump-traps you need to dodge to survive. But in essence, there are four decisions which will determine the bulk of your misery or happiness (okay, I'm not positive that one can truly be happy in graduate school, so let's say.... not-misery) in graduate school. Here they are, in order.
1. Make sure you really want to go to graduate school. This sounds silly, but I can't tell you how many students I ran in to who had not actually thought about this decision very much. Some did it because they were done with college and didn't really want to get a "real" job. Some did it because they did an undergraduate research project and kind of liked it. Some didn't know what else to do, and some like the idea of research and science but don't yet realize they hate the day-to-day. By my estimate, about half of these people drop out of graduate school, and the other half don't quit, usually for the same bad reasons they decided to come. Then they end up doing something totally different, usually after languishing for years. These days graduate schools are trying to limit entry of these people by requiring at least a year of post-undergraduate research, but sometimes that isn't enough. It's up to the individual to really understand what research is like; how political and grueling and repetitive and sometimes empty it can be, how much reading and writing and 'rithmetic is involved, etc etc. Make sure you understand these things, plus what kind of careers and pay is available to you if and when you get that degree. Make sure you are really happy with at least two of the options.
2. Choose the right department. Don't make this decision based purely on the subject matter. Know important things such as the nature of the preliminary (or qualifying) examination, class and teaching requirements, graduation requirements, reputation, pay, etc., all of which can sometimes vary widely from department to department. If you are an MD/PhD it is very important to learn how each department you're considering has treated mudphuds in the past, because believe me, this can vary wildly too. Often, if you like a particular primary investigator (PI, the person who would be your mentor/boss), and the project you are interested in is related to another field with a department which is a better fit for you, the PI will consider getting a joint appointment so you can have your cake and eat it, too.
3. Choose the right PI/lab. I've mentioned this in the past, but it is very, very, very important to find a PI whose style and personality is a good fit for you. For example, if you are a normal, balanced person, you would not want to work for a Yankees fan, since they all have strong tendencies towards evil. Don't work for a control freak if you are a control freak. Don't work for a slacker if you are a slacker. It's kind of like finding someone to marry, someone who will balance out all of your crazy parts, and whose crazy parts you can in turn balance out. Also it is better if you don't want to stab them with any handy object, blunt or sharp, after speaking with them for short periods of time. As I said, the other people in lab can be a good measure of this. Ask yourself: Do these people look despondent? Depressed? Are all the windows nailed down to prevent further suicide attempts? Are they angry? Drooling? Does their skin show signs of having seen even a few seconds of sunlight in the past year? These are important things to know. Also, ask about people who have recently left the lab. Did they do so in a tight, wrap-around coat and a white van, escorted by men with tranquilizer guns? Or did they move on in a normal span of time to a respectable position at a respectable institution (not the kind of institution the white van was going to)? Also: do you hate them?
4. Choose the right project. I learned this one the hard way. It might seem all shiny and flashy and awesome to start a new project, but this is the kiss of graduate school death in 95% of cases. Read that sentence again, because you, like dumb ol' me, will almost definitely be lured into this crap. You have to remember: It almost never works out for you. Who it works out for is the graduate student who follows you. And what are you, Mother Theresa? No, you will never graduate with that kind of altruistic attitude. What you need to find is some other poor Mother Theresa sap senior student who was tricked into starting such a project and is leaving and is ready to hand it to you with the four or five years of crappy, monotonous leg work all done. You might have to be co-author on the first paper (maybe), but believe me, it is almost always worth it. Also, the PI should be able to outline for you what sort of stuff will be included in your first paper in a decent amount of detail, and it should make sense to you and stand up to some hearty questions.
So there you are. The four main decision tree branches leading to graduate school success. Just try not to fall out of the tree, and you'll be all set.