Pipe is "measured by the ID"???.
I thought piping systems were designed by the ID for a specific flow and for MAOP by their schedule (wall thickness). All pipe of a specified NPS, ie. Nominal Pipe Size is of the same OD. All 4" NPS is going to be 4.5" OD.
This practice makes tap and die manufacturers to only have to have thread sizes for a few basic OUTSIDE diameters.
Here's a what if...
1/2" sch. 10s is   0.840 OD with a 0.083 wall thickness. Which gives a 0.757 ID.
3/4" sch. xxs has 1.050 OD with a 0.308 wall thickness. Which gives a 0.742 ID.
So if pipe is measured from the, ID then 3/4" pipe is smaller than 1/2"????

Also another glitch is that schedule 40 and "standard wall" pipe are not always the same wall thickness.

Shawn, take quick look at any steel shape chart to see what Giovanni is saying....been that way for a long time. The break is @ 14" on any of ANSI Pipe Schedule Charts that I have in the office.

I didn't know that You used the same system there.

On American engines We often used 1/8-27 NTP for oil pressure sensors.
Forign [Honda & Peugeot] engines I have owned use something similar but different, presumably metric that is about 26 threads per inch,
and I think tapered as there is no gasket.
Any idea what that thread or system is?

Actually, the Original Poster's Original Question was how to distinguish between Pipe and Tube.

And this ALWAYS becomes the contention point for all parties who participate in an answer: Do you measure pipe by the ID and tube by the OD?? 

Yes, AND  No.  As a large percentage of pipe is sch 40 and smaller than 12" (I didn't say all or most just a large percentage) the fastest way for most people to check the size is to measure the ID as they ORDERED 4" pipe and that has an ID that is VERY CLOSE to 4".  At the least it is easy to tell it from 3 1/2". 

There are so many different codes being worked to and so many different ways to call out the product that to get into a major disagreement about rather pipe is measured by the ID or not is not really the issue.  And even AISC sets up a practice for measuring by the OD X wall thickness with measurements to three decimal places.  This can apply to both pipe and tube. 

And if I am not mistaken, because of the vast differences in schedules and usage there are more and more industries calling out pipe by the OD and wall thickness instead of using the previously accepted practice of say, 5" sch 40 pipe.  More are calling it by 5.563X.258 pipe.  But not all.  So it becomes more of an issue when we see it on in a set of prints to ask for an RFI even if we think we understand what they want.  At least when talking structural application.  I can't tell you how many times I just happen to mention the size of a requested pipe only to find the engineer looking at me like a calf at a new gate.  I have to back up and make sure that the column he requested in 4" pipe was indeed 4" ID and 4.500 on the OD.  No wonder it wouldn't fit where he wanted it.  He wanted 4" OD!!  OR 3 1/2" pipe according to the manufacturers. 

If you want to know the difference you need to know the specs your material was manufactured to by requesting and comparing your product to the MTR for it's lot/heat number.  The particular ASTM number it is manufactured to will tell you rather it is tube or pipe. 

They really need to standardize the measuring practice so there is no room for confusion no matter what trade, industry, manufacturer, or Code you are working to.

One thing we can say for certain: Tubing Is Always Measured by the OD with the Wall Thickness.

A discussion on the difference between Pipe and Tube needs to get away from how PIPE is measured.  That is not the issue. 

And thank you for the ANSI number.  I had been trying to remember where I had seen a Standard that gave that information. 

Have a Great Day,  Brent

I'll give an example to explain better what I'm trying to say. In Australia you've long abandoned the British measures, so I'll use metric units.  

Let's suppose I've got to size a piping carrying 100 cubic meters an hour of water at ambient temperature. I'll adopt a velocity of 1,8 meters per second of water within the pipe.
100 cubic meters per hour is equal to 0,0278 cubic meters per second. 0,0278 cubic meters per second divided by 1,8 meters per second gives a cross sectional area of 0,0154 square meters. 0,0154 square meters corresponds to a diameter of 0,140 meters or 140 millimeters. This is of course the inside diameter.  

There's no pipe whose inside diameter is 140 millimeters, so I'll choose the next upper inside diameter, 154 millimeters, which is the inside diameter of a 6 inches, Schedule 40, nominal size pipe.
Now, I'll recalculate the velocity within the piping to be able to calculate the loss of head of the water flow.

The cross sectional area of the selected pipe (6 inches, Sch 40) is 0,0186 square meters, based of course on the inside diameter.
0,0278 cubic meters per second divided by 0,0186 square meters gives a velocity of 1,49 meters per second, which I'll use to calculate the loss of head.

All of the calculations have been made using the inside diameter.
Cordially
Giovanni S. Crisi

Shane/Giovanni,
In a sense you both are right. One is talking predominantly of manufacturing pipe the other design considerations.
And especially in power the manufacturing specifications for pipe diameter is often irrelevent. Heavy walls for steam service is often designed by ID regardless of ASME manufacturing specifications.
The design engineer will function just as Giovanni mentioned determining ID and minimum wall.
Min wall pipe is quite common in power. And it is always determined by ID since you cannot find manufactured wall thicknesses exactly as designed. Otherwise the number would be theoretically infinite.
ID is based upon the desired flow velocity, thickness upon temp and pressure. The OD will be whatever happens to be available at the time, within limits of course due to weight, restraint, cost, and support issues.