That study seems to only look at Amtrak trains, so it doesn't consider the large number of subway or urban light rail systems in use. Amtrak has relatively few passengers (along with many other problems). The NYC subway system, for example, is far, far more fuel efficient than a car. As is probably any moderately used metro service.

The article says underutilised, old and inefficient US Antrak trains are less efficient. Thsts quite believable, but isn't a general comment against trains.

From the article: "In the US: Cars are generally more efficient than trains when there are 2 or more people traveling."

That's rare. The vast majority of traffic is 1 person per car, which completely changes the results here to make it worse than the rest.

Sounds like trains are less efficient than cars maybe only in the US, because of lots of empty seats and/or less cars per passenger train:

>In the fuel efficiency information provided below you’ll see that trains are unexpectedly low, but this data can be a bit misleading because it’s from research that only studied trains in the US. There are two things to note:

>1) Americans don’t utilize trains as a mode of transportation very much in comparison with European and Asian countries. This brings fuel efficiency down because there are less people being transported per train moving.

>2)Trains have varying fuel efficiencies themselves. They are powered by either a normal combustion engine using diesel fuel or an electric motor. The efficiency of electric trains is highly variable because of the source of electricity. Electricity from a coal fired power plant will be much less efficient then renewable resources such as hydroelectric, wind, or solar.

Normalization on a per-mile basis misses the point. You build dense, transit-oriented cities specifically to reduce that denominator. Moving a person one mile has no intrinsic value whatsoever.

When you measure comfort per mile, cars are nearly infinitely superior to trains.