Blueprints

Start with downloading a Map Editor (I use WorldCraft) and run through the tutorial. That's where I started. Explore these VERY important and informative sites: The Forge and TF Console (links on left). You will need to learn a good Map Making editor (I do suggest Worldcraft)

You will need to add Team Fortress entities to Worldcraft's entity list using the Worldcraft FGD Patch. Or if you want an easier way (and haven't modified your Quake FGD file), just download my Quake FGD file and put it in your WorldCraft directory. I suggest saving your copy somewhere safe for a backup.
You will need to acquaint yourself with Team Fortress Entity functions with Robin Walker's TFORTMAP and TFENTREF.TXT. Download the both of those here. Also you can read about them here.
You will discover that building a map takes quite a bit of time. Iceland3 took at least 112 hours. Unholy Kingdoms took well over 200 hours.
You will need to learn how to keep your map from being laggy. Read this page!
You will run across all sorts of errors. Most were documented at QuakeLab, but that site is gone now.
You will wait through a VISing process which seems like it takes forever. Warships took 40 hours to VIS. Unholy Kingdoms only took 2.5.
You will need to have your map play-tested on a server.
You will receive criticism - some constructive and some painfully harsh - about your map.
You may discover that players abuse certain parts of your map. You may have to redesign it.

I don't mean to discourage new map makers by any means. But if this list looks impossible, perhaps try bouncing your idea off the heads of a few mapmakers to get them to make your map for you, or to at least find out in theory if your map will work. If they build it for you they will get the credit, but I should hope the mapmaker would add a line of credit for your idea in the map's TXT file.

Now for the good news: Creating a world in which many people will enter cannot be defined as anything less than the most rewarding experience of a lifetime.

Q: I want to make a Big, Wide Open...

The thing that some new mapmakers don't realize is, it can't be "BIG" or "WIDE OPEN." Okay, instead of saying "can't," howabout "it's going to take a lot of hills and other blockades." I learned this the hard way, so I'm writing this to save people some time and frustration. The Quake engine sees everything you see, and more - which can cause unnecessary lag to a map. Read this page to learn more about how to control lag. Placing burms, hillsides, doughnut holes, and other blockades will help quite a bit - but then it's hardly "wide open" anymore, is it?

Q: Why do you use Worldcraft?

I tried a couple others when I decided to try my own hand at mapmaking. That was in November of 1996, and I don't even remember their names. I was very impressed at how smoothly and easily the WC editor allows you to run around in your 3-D representation window. This means I didn't have to run my map in Quake quite as often. I've been hooked ever since. It's easy to incorporate TF entities, I can add in any settings for the TF entities right there in the editor. I can also edit > copy > paste from map to map, which is very handy. Among other features, like marquee-grouping. I registered it just because I thought Ben Morris deserved his money, and it was only thirty bucks. Even if there are better editors out there now, there's no way I'd switch. Learning new map editors isn't easy *or* fun, and it would seem like a horrible waste of time when I could be making maps, instead.

Q: Where do you get your map ideas?

Hmm, that is a good question.

Warships was Ambush (my husband's) idea. We needed a small map back then, when our server was *ahem* not busy all the time!

Gold Rush was sort of an accident. I had been playing around with a few mine shafts and a cart. Ambush said my cart was too small to carry coal. I retorted that it was large enough to carry gold. Then I thought how neat it would be if people went hunting for gold instead of the same old boring 'go get the flag' routine. The western theme just enhanced the idea, and I went to the planning stage right away.

Iceland was brought on by an intense desire to go cross-country skiing, and at the onset of Winter (when ICELAND was first conceived) seemed to be a good time for a map like that. I can't sit at Worldcraft and start building and hope a good map comes out of it, it never happens like that. The closest I came to that ever happening was Gold Rush. Usually when I come up with an idea, I let it roll around in my head from sometimes days to a few weeks while my mind works out the details.

Unholy Kingdoms was borne while I was toying with Iceland, trying to turn it into Greenland. I found one particular texture that just amazed me. I took the sniperloft area of Iceland and pasted it into a new map file, and played more with the texture. Then I went about making the castle's courtyard, set up some suits of armor, and did a poly surfaces count after a fast VIS just to see how ugly the numbers were. They were under 400! Bouyed by this information, I continued to build and build. I didn't care if it ever got released, at the time. Building this beautiful thing was not something I could stop myself from doing. I've never felt that way about a map, before. "This map has soul, and it definitely possessed me."

Q: What do you think are good ideas for a map?

Another good question. I don't by any means expect mapmakers everywhere to start building maps like I describe in this section but I suppose I can offer up some of my own findings here. Every time I see someone yell "this level sucks!" or "this level rocks!" on our server, I ask why. The players are, figuratively speaking, the buyers of our products. If enough of them like our maps, they will get server admins to put them on rotation, or they will go to a server that has that map. Server admins don't want empty servers!

The players seem to like these things:

Moving things. Trains are a big hit, especially complex ones. Platforms, doors, trigger_pushes, moving things.
Warps. Why are there hardly any warps in TF maps? I've been thinking how it's a nice way to keep enemies from ammo rooms, and it can also be used to cut down on r_speeds. If the ammo room doesn't connect to anything, a MapMaker can just go gung-ho in creating a very spectacular place to respawn to. It would be a safe place to type things like "Flag is outside our base!" or even "Hi Dyer!" as well. :}
Things they can make work, or disable. The cannons in Warships. The way The Rock's mine shaft can be fixed. The way Engineers can operate things in the 2Tech series. Disabling the elevators in Well6 (engineers can do this by building things half on and half off the elevators, ask one to show you).
New ways of scoring. Gold Rush (bringing gold to your bank's vault), Canalzon, The Border, Hunted President. The only problem with maps like Hunted is that the "guards" can kill the President, which makes the map woesome when people don't know how to play it. TF MapMakers are hoping TF2 will solve problems like this that MapMakers face.
Several ways into the key. Someone told me 3 ways in and 2 ways out are nice. Making a hole you can drop down into but never get back up would be a way to do this (Oppose1), or putting a trigger_push that only allows a player to go one direction (Oppose1 again). Three ways in allows the offense a good chance, and 2 ways out give the defense a good chance, too.
Make them work together at something besides scoring. For example, Gold Rush has banks in which the players work together to "buy" bank alarms. The first level of the bank's alarm sets off an audible alarm. The second causes a little pain to the enemy when he enters and leaves the bank (and the alarm goes off). The third and final level gives everyone on that bank's team Quad damage whenever an enemy enters or leaves the bank (as well as the first two alarms). The players are so busy bringing gold to upgrade their bank's alarm, its like a mad rush to the gold mines. Few shots are fired, and they don't much bother with scoring until their bank has been upgraded. Since the enemy team can rob their bank and score, they want protection. And besides, who doesn't love Quad damage? :} The fate of the game rests upon who's bank has been upgraded and to what level.
Architecture. Often a map becomes a favorite just because of it's beauty, but us TF MapMakers have to be so careful of the r_speeds. We need to counter one with the other.
Water ways in to the base is a big hit, in case the entire enemy team consists only of snipers.
Levels that allow each class to play their best. Engineers need places to put sentries. Snipers need places with shadows to hang out. In Iceland3 I've given snipers a team-only and class-only bag that restocks them slowly on the sniper ledge so they don't have to run back for ammo, health, or armor. They do like to be alone in their world. I've also given them illusionary burms in the snowfield to hide in. HWGuys like long-range line-of-sight so they can see an enemy coming and have time to spin up. Demomen like to have something to detpack. You get the idea. We won't be putting Sniper War on our server, sorry.

Q: I have an idea for a map. Will you make it for me?

I get this quite a bit. The answer is another question: What is your idea? First and foremost, I can point out things in theory that might not work. Secondly, I'm a person like any other, and if the project doesn't appeal to me, I won't want to make it. But, bouncing ideas off the brains of map makers is a good start! You can at least find out if, in theory, the map will work. You can then decide to try your best at building it yourself - or allow the map maker to build it, and post credit from the idea somewhere in the map or text file. He doesn't have to, however. There is no copyright protection for an idea! It may be that, after you expose your idea, the map maker tells you 'no,' and then builds it anyway - professing the idea was his the whole time. Scary, yes. But if he built it, he suffered a little, trust me. :}