I've had at least two "cold weather" tanks so far, which were 38.5 and 39.1, so I'm down 2-3 MPG over the same type of driving done in the summer. I'm now at 4500 miles or so, with 50% oil life left on the original oil.

 

REcently proven, ethanol blends up to 65% on non-flex gasoline engines actually boost mileage.

I think it has to be many other above things you mention. I just want to add one, which has nothing to do with winter, which is the new car is breaking in and the oil is just getting dirtier.
For me personally only after my 2nd oil change it got better. I had the same situation where the fit started off amazing 35mpg then dropped to 32. But it went back up.

 

Have you got a link for the proof? I'd like to take a read of that myself.

Personally, I've seen a drop in mileage recently as the temps have dropped a bit, too. My commute is about 5 miles each way, so probably half of each trip the engine is still warming up. Plus, running the defrost is a must, so the AC is engaged. Still, back when I got the car in October, it was hot enough to run the AC and I got a mile or two per gallon more than the 27.6 I got on the last tank. 30.4mpg all city on the very first tank, with 27 last and 28.9 the tank before that. Granted, I don't baby the car, but I'm not hard on her either. Usually shift about 3K, and in to 5th by 40mph or so in town.

 

Green Car Congress: Study Finds Certain Ethanol Blends Can Provide Better Fuel Economy Than Gasoline

Hopefully this means something useful.

Remember cold air affects tire pressure. If your commute is far (as in my case, where I drive 35 miles one way), since the tires would heat up for most of the ride, the mileage won't suffer. If you have short commutes (like your five miles), youre driving mostly on low pressure tires, hurting the mileage.

 

Thanks for the link! Now to find some time for a little recreational reading...

I'm a little more anal than most on tire pressure, and I also like to run mine around 38psi (36psi on our Pilot) to help protect the wheels from the lousy roads around here. Still, 38psi at 90* = something less at 40*, so it's a moving target this time of year. I keep an air tank right inside the garage door just in case.

 

Try this:

Find a long stretch of open flat road that you can drive at highway speed. Make sure the engine is completely warmed up.
Drive the car with the throttle held in a constraint position (do not use the cruse control) for about 2 minutes. Remember to hold the throttle in constraint position - let the car speed do what ever it's going to do with the hills in the road.

What the engine control module (ecu) will do during this period is read the oxygen sensor in a closed-loop mode and will recalibrate the mapping for the fuel injectors and timing.

I do this periodically during my highway commute to work when ever there is significant change in the weather. I have never gotten below 37mpg since driving my 5sp Manuel FIT for 28k miles. I drive about 70% highway. I got this Idea from reading the owners manual in the section talking about passing emission testing. I would like to see if this technique works for other FIT drivers out there. Please post up with you results.

 

First, Gordio, thanks for that link! Interesting reading, but probably not very useful until we can vary the blend of ethanol at the pump to be ideal for whatever car we're filling. I don't see that as being a technical issue, but more of an R&D plus supply issue.

I tried the steady-throttle trick, and I can't say it did anything for mileage. Two fillups back, I got the worst mileage yet - about 22mpg, all city and cold weather. Then I took a nearly 200-mile road trip, and managed nearly 37, the best yet. Go fig.

And it looks like 38psi @ 90* = 30psi @ 40*, so the tires need a bit of air. Unfortunately, I didn't air up before making my big mileage run. I wonder if I could have hit 40mpg if the tires were at 38psi like I try to keep them?

Hmmm...