Europe's plans to send a robotic rover to the Red Planet in 2013 face a critical review this week.
A top-level panel will meet in Paris to choose a single design concept for the mission and determine whether the ambitious proposals are affordable.
The European Space Agency delegates are being asked to approve an upgraded - and much more expensive - vision than the one originally outlined.
This would see ExoMars travel with an orbiter to relay data back to Earth.
Enhancing the mission in this way would free the rover from having to rely on American spacecraft for communications with home.
It would require the whole package be launched on a bigger, more powerful rocket; but this also means the robotic vehicle could traverse the Martian surface with additional, or heavier, instruments to search for signs of life.
"What we've been looking for is a concept that will maximise the scientific return," Bruno Gardini, a member of the ExoMars project team, told BBC News.
The team's recommendations will be subject to a review on Tuesday before being handed to the Programme Board for Human Spaceflight, Microgravity and Exploration.
The latter should settle on a final design concept and assess the need to increase ExoMars' 650m-euro budget. Many tens of millions of euros may be needed. A decision is expected on Thursday.
For months, scientific and industrial groups across Europe have been developing a number of concepts in parallel.
The early "baseline option" called for a rover to be launched on a Soyuz-Fregat booster and landed on Mars using the bouncing gas bags employed so successful by the current US vehicles, Opportunity and Spirit.
In this baseline configuration, the European rover would have the capacity to carry 8kg - perhaps a little more - of instrumentation, including a drill or "mole" for burrowing beneath the Martian soil.
A Geophysics/Environment Package (GEP) would also be landed with the rover. This fixed station, once placed on the planet's surface, could sense "Marsquakes" and monitor the weather.
But the closer this concept has been investigated, the more limiting its capabilities seem. The bouncing bags put tight constraints on the volume of space into which the rover must be stowed.
This in turn restricts the maximum mass that can be given over to instruments and makes it virtually impossible to carry a meaningful GEP payload.
Disquiet has been expressed, too, at the configuration's intention to use American orbiting spacecraft at Mars to relay commands and data. If the US were to experience a spacecraft failure - as happened late last year - communication could still be possible but the ExoMars mission would be severely compromised.
It is for such reasons that design teams have been considering a beefed-up concept. This would launch the rover on a heavy-lift rocket - an Ariane 5 or Proton.
A bigger boost at launch would provide extra mass capacity for ExoMars to travel with its own orbiter, which itself could do remote sensing of the planet.
This concept also envisages using novel "vented airbags" to cushion the rover's landing. The military technology inflates like pillows under the vehicle and can support heavier payloads.
"For the rover's scientific instruments, the option of a bigger launcher gives us something in the range of 16kg, which is twice as much at the American rovers," explained Mr Gardini.
"In this case, we would also be able to carry a geophysical package - and we are still doing the final calculations - probably in the range 20-30kg."
EADS-Astrium has been developing the rover's chassis at its UK base in Stevenage. It has built a testbed to examine the vehicle's locomotive capability, and to try out the autonomous navigation software that will eventually guide ExoMars over the Red Planet's rock-strewn landscape.
Like all the design groups, Astrium has had to grapple with the different mission scenarios and what they would mean for its segment of the project.
The company's ExoMars group leader, Mark Roe, is convinced a heavy-lift option is the most appropriate way forward.
"First of all, it would be good news because we would then get maximum science benefit. And that's good news for Britain because we have people here who are well positioned to be principal investigators in a significant number of instruments," he explained.
"It's still going to be very challenging for the launch mass. You feed that back and you find the mass of the vehicle, which is now about 190kg to accommodate 16.5kg of payload. The smaller launcher will not give us the mission that we really want."
Some commentators are worried that this week's events could lead to a "go, no-go" decision.
If it is felt that the heavy-lift option is simply too expensive and the current baseline design is viewed as delivering insufficient science, then ExoMars could be shelved according to this viewpoint.
But few would be prepared to contemplate such an outcome. The rover mission is the first step in an ambitious programme of Solar System exploration, known as Aurora, which could eventually see European astronauts walking on Mars.
To abandon ExoMars now would make a mockery of the programme and raise serious questions about Europe's space capability.
It would also knock back a generation of scientists who have pinned their hopes on the Aurora roadmap to deliver a Red Planet sample return mission - a mission to bring back pieces of Mars rock for study in Earth laboratories, where the full panoply of modern analytical technologies can be deployed.
"The idea of sample return missions is generally becoming the thing people really want to do in the planetary science community," observed Dr Ralph Cordey at Astrium.
"Japan has had a go with its little Hayabusa probe and in Europe we are very much hoping there will be a mission to an asteroid to do a sample return.
"There is a limit to what you can do in situ, however many instruments you take to these bodies. Bringing material back to Earth would represent a huge step change in what you could achieve."
This week's meetings have to deliver clarity. Time is running short. The launch of ExoMars has already been put back from 2009 to 2013, and industrial contracts need to be issued soon to keep the present schedule on track. If a single technical baseline is adopted this week, design teams can then refine the concept.
"It would enable us to start to optimise it, which hopefully will bring some mass benefit which, hopefully, we can then feed back and, maybe, even get a slightly bigger payload. That would be terrific," said Mark Roe.