In another question - Proportions of a space-craft as compared to fuel - I explored the Tsiolkovsky rocket equation and the tyranny thereof. In the related NASA link, Don Pettit explains that, even using the Hohmann transfer orbit between Earth and Mars which is the most fuel-efficient, it takes about as much change in velocity to break LEO and head out toward Mars as it does to go from the launchpad to LEO. Where does that energy come from? Well, it comes from shoving superhot gases out your tailpipe.
So, given a spaceship roughly the mass of the Space Shuttle orbiter (just as an example; the Mars vehicle may be larger, maybe smaller), which is 110 tonnes plus up to 24 tonnes cargo, to get that into LEO we normally burn about 725 tonnes of LH2/LOX fuel through the main engines, plus about a thousand tonnes of perchlorate fuel in the SRBs. Gross liftoff weight of the Shuttle is 2 thousand tonnes, of which the orbiter vehicle and cargo is a maximum of 6.7% of gross liftoff mass. There's some mass inherent in the support structure of the tank and SRBs than is accounted for here, but not nearly as much as you'd think; the SLWT external tank, for instance, has an empty mass of only 3.7% of its loaded mass.
Once the Shuttle's in LEO, that's typically it; it carries a small amount of fuel for its deorbit burn and for maneuvering, but the main engines are cold for the entire remainder of the trip. However, we're now talking about getting that mass out of LEO and on a transfer orbit to Mars. That actually requires about the same delta-V as getting to LEO and therefore about the same amount of fuel (we have to burn a little more fuel to get out of the atmosphere because of drag; the Tsiolkovsky equation assumes an "ideal" - drag-less - rocket). So, take that 135 tonnes, strap another 1725 tonnes of fuel and maybe 140 tonnes of support structure on it, and blast off again from LEO to parts unknown.
... But wait, where'd that fuel come from? We have to get the fuel out of Earth's gravity well. That requires us to lift 1725 tonnes of payload into LEO on subsequent launches in order to "refuel" our orbiting craft. Assuming we can use a Shuttle-derived vehicle, like the Space Launch System, to do that with a similar per-launch total payload as launching the Shuttle (it will eventually be possible; SLS Block II is spec'ed to carry 130 tonnes of payload to LEO and we can probably improve on that), and also assuming the spacecraft we launch can keep its fuel tank instead of jettisoning it as the Shuttle does, it would take about 13 launches (those launches consisting of basically a big fuel tank on top of a bigger fuel tank) to get enough fuel into orbit to send our craft on its way to Mars. Each of those launches would burn 1725 tonnes of fuel to get 135 tonnes into orbit, for a total fuel cost of 22,425 tonnes.
... But wait, we want to be able to get back from Mars. Well, that requires a similar delta-V as getting there in the first place. So, we need 1725 tonnes of fuel to get our 135-tonne spacecraft out of Mars orbit and back down to Earth. How does that fuel get to Mars? It rides with the spacecraft. And that means that we need more than 1725 tonnes of fuel to break Earth orbit. In fact, it needs the same 22,425 tonnes that we calculated we'd need in order to get the 1725 tonnes of fuel from Earth to LEO, which will now be used to break Mars orbit. This 22,425 tonnes will be used to break Earth orbit and get the 1860 tonnes of vehicle and return fuel out to Mars.
And that fuel now needs to be lifted out of Earth's gravity into LEO so it can be strapped on the back of our much-larger rocket. To lift 22,425 tonnes of fuel, 135 tonnes at a time, up to LEO would require 166 more launches, in addition to the 13 needed for the return fuel, plus the one launch for the actual vehicle, for a total of 180 launches from Earth's surface to LEO in order to get this ship in orbit and fueled for its departure. In other words, we'd need more SLS launches carrying just fuel than total combined Saturn V and Shuttle launches (which are the only two vehicles we've ever launched with the lift capacity even close to getting the job done). Each of those launches needs 1725 tonnes of fuel, for 310,500 tonnes of fuel burned just getting the vehicle and fuel into LEO. Then the ship itself will burn 22,425 tonnes of fuel getting to Mars, then 1725 tonnes getting back, for a total fuel expenditure of about 335,000 tonnes. So, fuel costs aren't too bad; for LH2/LOX, at an 11%-89% mixture and today's prices (\$5.50/kg LH2, \$.20/kg LOX), we're looking at around \$250 million in raw fuel cost, plus off-gas losses (liquid hydrogen and oxygen don't just sit around in liquid form at room temp).
However, total launch costs are a big deal. The SLS, if it hits its cost goals, will be about half a billion dollars a launch. 180 launches to get the vehicle and fuel into LEO represents a cost of about \$90 billion just to get the materiel into space. Actually designing and building what we're launching could exceed a trillion dollars, given that it can not fail; if the crew has a "problem" halfway out to Mars, like the one Apollo 13 did, the chances of them getting back to Earth safely are nil.