Maglev would work very well on the moon for launching materials from the surface into space. Between the low gravity of the moon resulting in lower orbital and escape velocities (roughly 4,000 and 6,000 MPH respectively) and no atmosphere, it would be much easier to use maglev on the moon than on the Earth or Mars. You wouldn't use it for delicate things like people but for bulk items like soil, water, or ice. This is an old idea, dating at least as far back as Gerard O'Neil's proposals of the late 1970s and early 1980s. The biggest problem with maglev is the expense and difficulty of building the launch rails and power systems. Once built, the cost per pound of launching something is extremely low. A collection of reference links is available here.

However, at least in the near term, maglev isn't required. There is another idea for low cost launches from the moon proposed by some people from Lockheed and the Department of Energy. It's called the steam rocket. If (and this is a fundamental if for all moon infrastructure ideas) harvestable ice does exist on the moon, you can use it directly as a propellant by heating it with a simple nuclear reactor to produce steam, thereby producing thrust. As the abstract states:

These cost reductions depend (exponentially) on the mission delta-v requirements being less than about 6 km/s, or about 3 times the specific velocity of steam rockets (2 km/s, from Isp 200 sec). Such missions include: from the Lunar Surface to Low Lunar Orbit, (LLO), from LLO to Lunar Escape, from Low Earth Orbit (LEO) to Geosynchronous Orbit (GEO), from LEO to Earth Escape, from LEO to Mars Transfer Orbit, from LLO to GEO, missions returning payloads from about 10% of the periodic comets using propulsive capture to orbits around Earth itself, and fast, 100 day missions from Lunar Escape to Mars.

Such a rocket system could be used in Earth orbit to power a reusable space tug, flying between LEO and GEO to both boost and service satellites. This would greatly lower launch costs because you wouldn't have to launch the stages needed for putting the satellite into GTO and then performing the apogee kick to lower inclination and circularize the orbit. A large 5 ton satellite bound for GEO could be launched on an inexpensive booster like a Delta II instead of requiring a really big and expensive booster like the Ariane V, Titan IV, or the heavy versions of the new EELVs. A Delta II rocket goes for about $20 million (although some versions with more solid fueled strapons go for quite a bit more), while a more powerful booster can easily cost over $100 million to launch the same payload to GTO. The Titan IV Centaur costs upwards of $400 million per launch.

The idea is that you launch the payload into a compatible orbit with the tug. Using standard attachment interfaces, the tug would dock with the payload and then boost it to GTO. Once in place, the tug would circularize the orbit and detach from the payload. Afterwards, the tug would do a retro burn to drop the perigee down to 100 miles or so and then use aerobraking to lower the orbit back down to rendevous levels. You could do the same thing launching probes for the outer planets. The tug would boost the payload to escape velocity, then detach and do a burn to come back to the Earth for the next mission.

Such a rocket could send the ice harvested from the moon (assuming it exists) to Earth orbit. There, could be used to either directly fuel other steam rockets or broken down into hydrogen and oxygen to both provide propellants and consumables for manned missions needing higher Isps. This is a big part about what I mean by space infrastructure. To the greatest extent possible, materials used in space would be harvested from space resources. This would include materials harvested from the moon and of course propellants produced locally on Mars. The only things that you would launch from Earth would be the things that you couldn't produce in space, such as trained crews, precision components, electronics systems, and the like. It'll be a long, long time before the facilities to do IC fabrication exist in space. The same is true for mills to produce aerospace quality alloys, build satellites and spacecraft, and grow, educate, and train the crews.

Launching anything from the Earth requires a lot of energy, therefore it's expensive. It's probably going to remain that way for a long time, if for no other reason than the fact that the companies who'd be most capable of producing lower cost boosters don't have much financial incentive to do so. Why spend billions developing a system that will compete and undercut the market for other systems that you've already spent billions developing? It'd be like a luxury car maker producing a lower priced car that does everything their expensive cars can do and just as well. It doesn't make financial sense. That being the case, developing an infrastructure in space that uses to the greatest extent possible resources from space is the best way to ultimately lower mission costs and to open up the solar system.

For a mission to Mars, the spacecraft itself would still be built on the Earth. It might be launched in modules and assembled in LEO like the ISS or perhaps launched in a single piece on a heavy lift booster derived from Shuttle technology (or a revived Energia). The many tons of propellants and most other consumables needed for the mission could be provided from lunar resources, eliminating many expensive supply launches from the Earth. The crew would be sent up on something like the proposed CEV.

Of course, all of this depends on whether harvestable ice exists on the moon. We need to send an unmanned landers to the lunar poles to answer this question as soon as possible.