Drove from Boston to Rochester. The way to Rochester is uphill, from sea level to 1724 feet - which doesn't sound like much but is the highest point on Rt. 90 until S. Dakota. The way from Boston was very rainy and my experience is that knocks a couple of miles off the count, particularly at higher speeds. The way back was clear. I forget the exact calculated - not computer - mileage to Rochester because the weather was bad but it was somewhere around 44.5 to 46. The way back I calculated 51.5, with a computer reading of over 54. The average speed was 67. Lower on the way from Boston because we took a side trip, but a true 67 all the way to Boston. This shows, with uncertainty from the rain, a difference of 5-6 miles, of which 2 to 3 or so is rain. So a difference of roughly 3 to 4 miles for uphill over downhill over a long distance. I also tested the rolling resistance several times by coasting downhill next to vehicles ranging up to large trucks. I could most often see if the other guy's brakes were on. The Prius rolls faster than almost anything. In one case, it rolled much faster than an SUV towing a heavy boat, which shows that lower wind resistance can beat gravity, as though we were using gas and they were using brakes. It rolled about the same as a full size tractor trailer downhill, though of course I don't know how much the truck was loaded. That impressed me.
My daily commute is about 35 miles each way, with an elevation change of 412 feet (my house is higher than my work). The rise and fall over distance is negligible, but discounting other factors, I lose a couple miles per gallon on the ride home versus on the way in. Typically the computed mpg in (downhill) is about 54, but on the way home, about 51 - 50. With your drive, as with my commute, I am sure it was a matter of going up and down hills of increasing height over the distance one way, and decreasing height the other way. I think it is this factor that has the greatest effect on the mileage.
With respect to MPG uphill vs. downhill, my estimate is that my Prius consumes about an extra gallon per 10,000 feet of climb, depending on cargo load. It recovers the difference on the way down unless the hill is steep enough to require braking. Ken from Japan has a bit lower estimate, but he and his car are lighter. As for rolling resistance while coasting downhill, remember that the other cars also have significant engine compression causing additional drag. In comparison, the Prius engine drag is tiny.
Everything falls (or goes downhill) at the same speed regardless of weight (or mass), except for the effects of friction. So a Prius would go faster than anything of any weight (more or less) that had more drag. This was proven a long time ago by some Italian guy who dropped things off the Tower of Pisa. This was also tried on the moon and of course it worked there too (they dropped a ball and a feather and they hit the ground at the same time). You can see more information on wikipedia if you look up "Leaning Tower of Pisa experiment". For some reason this forum won't let me post the link.
Generally speaking, the prevailing wind on that route is from west to east, which could be a major factor re: mpg. Do you remember what the wind was doing each way?
