Aaaragh!
Lost my post...damn computers.
Anyway, starting over...
GMing is learned via experience. If you suck at it, chances are you'll get better. Almost all new GMs suck, and those that don't are merely mediocre.
It takes time to get it right, and it takes good players.
Uncle Figgy (do a google search) wrote a pretty useful primer on the subject, and Unknown Armies and Call of Cthulhu D20 have the best GM sections I've encountered in an RPG, bar none. Particularly UA- if there's a single game that's influenced how I run other games, it's Unknown Armies.
I'm a Ninja GM now, skilled in all the deadly arts, and I can give you some rules of thumb.
1. Never just say No." Say "Yes, but..." or "Yes...if..." or "No, but how about..." or "Maybe...if you can..." "Go ahead and try..." If you have to flat-out say "No", explain why- good players are not selfish players, and if you have a good reason, they'll not have a problem.
2. Don't worry too much about it. If your players do not complain, or do not raise suggestions, then you're almost certainly doing a fine job. And if you aren't, you owe them nothing. They are just as responsible for giving you feedback as you are for giving them a good time. How can you provide fun if you don't know what "fun" means? A GM filled with self-doubt cannot have a good time himself, and a GM who isnt' having fun almost certainly ends up a bad GM, and even if not...why are you GMing if you don't like it?
3. Bad players are not worth it. I'm going to say that one again. Bad players are not worth it. One more time: Time and energy you expend keeping a jerk happy is time and energy you aren't spending making good players even more happy. If they in any way make the game less enjoyable for you and the other players, kick them out. Kick them out without hesitation, without remorse. No game can succeed without quality players.* Of course, warnings, talking about it, etc, before a kick-out are usually warranted, but I find that good players are never consistantly disruptive, even unintentionally.
*: By "good", I simply mean players who are considerate of the rest of the group. I do not refer to their ability to play a character, which is a skill, like that of a GM's, that takes time to develop.
4. "Railroading" is an overused term. It does not refer to a campaign premise or restrictions on character creation. If you want everyone to be cops in 1970's Las Vegas without supernatural power, that is not railroading. There are implicit contracts to games and explicit ones for campaigns; if they agree to play Call of Cthulhu, then they agree to all that entails, and if they agree to be Arctic Explorers, then they agree to certain limitations on things.
"Railroading" really only refers to a GM trying to impose certain choices on a character's actions. Your players have absolute control over their character's free will; you do not. If they aren't going to murder the prisoner, and all your plans were built around that prisoner's murder, then you need new plans.
However, there is a trick to getting around this: asking them. "Say, Bob, can your character be mad at Betty and storm out of the apartment? I want to kick-start a kidnapping storyline and have Bob feel guilty for not protecting her." Good players almost always either say "Sure!" or come up with a better idea. If they say no, they'll almost always have a good reason, and you should get new plans.
5. Corollary to that last point: Hat Tricks are an excellent GMing tool. Basically, you take a single player into your confidence for a specific set of events, like being demonically possessed. Eg: "Psst...Bob, your character is going to be possessed next week. Want to play the demon?" Players LOVE this stuff. You can't do it all that often, and don't do it only with the same player again and again, but it allows you to pull off stories you might not have been able to do otherwise.
6. No plan survives contact with players. Part of the fun of GMing is the "Herding Cats" aspect- the improvising to the sheer chaos of your players. They'll surprise you, and this is a good thing. As such, have only skeletons of plots and ideas for good scenes. Drop things at will- if you can't get something to happen, it doesn't happen. Try something else.
I usually have basic notes on scenarios, and the notes got sparser and sparser as time went on. You'll need a bit more if you have players who aren't proactive, but I find that once you get the game started, they're the ones who end up finishing it. Generally you need NPCs, their agendas, maybe notes on locations, and a general idea of "what kind" of adventure you want. Light-hearted? Dark? Dramatic?
Have an idea for "cool stuff to happen", but you'll more often than not end up with something entirely different.
Essentially, RPGs are a kind of collaborative storytelling exercise, and one of the biggest mistakes you can make is having the story done beforehand. That isn't the point- the story is there at the end, not at the beginning.