Econbrowser

Slowing down really works.
300 mile trip, first 20 heavy Sacramento traffic, last 280 miles on lightly congested freeway,
1991 Volvo station wagon with good cruise control.
Trip out, 70 mph, 21 mpg.
Trip back, 62 mph, 26 mpg.
Irony is that the trip our began with 20 miles of heavy traffic leaving Sacramento. The return arrived Sacto during low traffic time of day, was 10 minutes quicker, door to door.

Thanks for linking!

Ironman, someone raised the possibility that accelerating fast may not be more efficient for automatic cars. It is plausible, but unlikely, that they do not shift at the appropriate times to take advantage of best engine operating speeds. I e-mailed Dr. Dougherty from the NY Times article that I linked before I left work yesterday. Hopefully he received the e-mail and will confirm accelerating is faster for automatics.

... confirm accelerating faster is more efficient for automatics too.

Thanks for following up Aaron. I wouldn't have expected a difference for automatics, although I would anticipate they follow a different acceleration/fuel-burn profile.

I do suspect that aggressive acceleration with an automatic transmission is worse than that for a manual transmission. When you floor the accelerator in an automatic, the transmission will downshift, and while that gives you more power, it also wrecks any chance you have of an efficient acceleration.

To double the speed of a car requires four times the horsepower due to turbulent airflow. This calculator doesn't seem credible.

Robert, (1) overcoming the friction from air turbulence is not the limiting factor at low speeds; (2) the engine is more efficient at higher speeds; (3) these are based on empirical measurements rather than your theory.