My car sometimes starts badly, then sputters and idles poorly! Is there a fix?

This is likely the infamous 'hot start' problem. This question pops up every year in the springtime, when many DSMs 'suddenly' exhibit this problem.

Perhaps better named as a 'warm start' problem, this difficulty often surfaces after a car has been driven, then parked for a relatively short period of time. Upon restart, the warm engine (not fully cooled down from the previous drive) appears reluctant to crank over. When it does catch, idle is often in the 300-500 RPM range, with engine shaking, sputtering, reluctance to rev up, and sometimes stalling. Holding the accelerator down until the engine smooths out often 'solves' the problem, but sometimes the engine will not rev up at all. Often the problem will correct itself with no intervention by the driver.

Please note that an inability to crank over or failure to actually start the engine are not related to the 'hot start' problem. The problem referred to by that name relates only to bad idling after starting.

This problem appears to pop up on every year of DSM if the conditions are right. Dealerships are often completely unable to diagnose or even replicate the problem.

First diagnosed in the 1990 year, Mutsibishi developed a 'kludge box' - an add-on ECU modification - designed to fix the problem, and released TSBs #18-08-91 and 18-55-91 describing it. Other model years have no such box available, nor are there any TSBs. Since TSBs are not warranties, 1990 owners may still be out of luck. There are no reports that the kludge box was effective anyway, but at least 1990 owners have a definite 'fix' to try out.

Later 1G owners have reported a bewildering array of 'fixes' to this problem. Owners have reported hot start problems that they have attributed to many different components. Fortunately, Jim McKenna put together a nice Hot Start FAQ  that should help you diagnose the problem. However, since the problem can arise from a large number of component failures, you may end up systematically going through your engine to find the problem component.

2G cars also have this problem, which is puzzling since they are largely different from their 1G counterparts. There has been speculation that there is some kind of flaw or design error in the 2G engine or ECU software that allows this problem to occur. In other cases, the ECUs were able to flag a specific error that helped the owners track down the errant sensor that was causing the difficulty. In any case, the fixes described in the Hot Start FAQ  may still help out 2G owners. Other fixes include changing thermostats, switching brands of gas, and general fuel and ignition systems troubleshooting.

It is possible, however, that some (or all) of these problems are nothing more than good old-fashioned vapor lock. Gasoline blends vary according to climate, and winter fuels have more volatility than summer fuels. (In other words, winter gas vaporizes better.) In warm temperatures winter gas might vaporize more than it should. Since fuel pumps cannot pump vapor, the engine does not get the fuel it requires to idle properly and it ends up stumbling. Eventually the problem cures itself once the vapor is cleared from the fuel system. Presumably it will go away in colder temperatures or once summer fuels become available.

The Last Word: No real consistent fix has ever been discovered for this problem. Personally, I think it's a combination of two problems: an ECU coding bug, where the ECU gets "confused" because the temperature and other sensor inputs do not correspond to the actual engine condition on a "warm start", and good old-fashioned vapor lock in fluctuating weather conditions. YMMV.

If the hot start fixes page is a 404, you can view the answer here: How do I fix the hot start problem?