Pipe threads are tapered and all the teflon tape or the liquid do is serve as a lubricant. As you tighten the fitting it seals itself.

Not all will seal - see this quote from DirectMaterial.com :

"NPT threads are designed to be used with a sealing compound such as Teflon tape to create a leak-free seal, whereas NPTF threads form a mechanical seal by crushing the entire thread form."

Standard coarse NPT threads are not malleable enough to seal against each other. And pipe threads are cut with dies, so many have too rough a surface to seal reliably (top quality threads are rolled, not cut, which is part of the high cost of AN hardware). Tape may be necessary even on some NPTF fittings. The lubrication role is also important, but it makes overtightening easier - so care is necessary to avoid breakage or distortion of lines & fittings.

Taper pipe fittings are intended to be tightened no more than 5 complete turns max and new fittings are usually fully tight at ~3 1/2 turns.

In regards to Teflon tape this stuff has been banned in nuclear power plants for years. The tape is tricky to apply properly as has been mentioned here. The other problem not previously addressed is if the initial wrapping is not perfect there is the high probability of small shreds of tape being produced that end up in the process stream.

There are paste types of pipe joint sealants that are far superior to teflon tape and easier to work with.

None of these products will reliably seal up a loose connection. The fit-up should be past hand tight and require a minimum of some minor wrench use.
Lots of info and choices here: http://ows.rectorseal.com/Pipe-Thread-Sealants.php Real USA products.

The Loctite 567 is an excellent product. As EricS said keep all sealers off the first one or two threads and do not over apply.

I do not recommend this but at a Loctite seminar for maintenance mechanics a fitting was over tightened and then very slightly loosened to put it in alignment with another tube and connected, the fitting did not leak even under hydraulic equipment pressures. I also witnessed this same thing done on steam fittings without creating leaks. Things like that can save your bacon when doing repairs in tight spots.

ETA: Product number corrected. Error on my part!

bettysnephew wrote:The Loctite 545 is an excellent product. As EricS said keep all sealers off the first one or two threads and do not over apply.

I do not recommend this but at a Loctite seminar for maintenance mechanics a fitting was over tightened and then very slightly loosened to put it in alignment with another tube and connected, the fitting did not leak even under hydraulic equipment pressures. I also witnessed this same thing done on steam fittings without creating leaks. Things like that can save your bacon when doing repairs in tight spots.

This only works with sealants that remain slightly flexible. Some fully harden and do not permit minor repositioning once applied & starting to set. Confirm from the manufacturer that what you use is OK for how you use it if the container or insert doesn't tell you. Read all instructions before use, every time you use something new to you or that you don't use often. I've made some big mistakes by assuming I knew more than I did. The older I get, the happier I am to find confirmation that I'm doing the right thing before I do it

I agree with Bluesman regarding knowing your sealants' characteristics.
The 567 does remain flexible even long after assembly.
Again another thing I do not recommend but have seen done in actual situations was when the mechanic did not have the tube of sealer handy and time was critical due to food product parameters, was to just reassemble the parts with the original 567 on the threads. It didn't leak and was repaired properly at the next shutdown/cleanup period. This was done in a possible cataclysmic situation as certain food products we made were extremely time and temperature sensitive and huge batches of thousands of pounds/dollars would be lost if overcooked. Not only the dollars thing but some of the products if overcooked would turn to solids that literally had to be broken into pieces by hand with hammers and chisels to remove after the machine was disassembled to expose it for cleaning. The sugars in the product would turn to rock candy like substances only much harder and burnt. After chipping out this hardened product the machine had to be totally sanitized before reentering production. We were willing to accept a small bit of leakage at the fitting versus losing an entire production batch or creating such a cleaning nightmare. The fitting did not leak a drop and normal production resumed after pumping off and offaling the product that was exposed in the repaired area.
545 is a paste when it comes out of the tube and never actually seems to fully harden in any amount of time that I encountered its use. It does get gummier but not solid.

bettysnephew wrote: 545 is a paste when it comes out of the tube and never actually seems to fully harden in any amount of time that I encountered its use. It does get gummier but not solid.

Maybe you are thinking of a different Loctite product. I have a bottle of the 545 in house & have used it several times recently. It is a liquid, not a paste although it does start setting up pretty quickly once exposed to the air. Directions call for a minimum 12 hour cure time with 24 hours ideal.

Lots of info on Loctite products here: /downloads/

OOOOPS!! Sorry,
The product I was referring to and using is Loctite 567 not 545. Mea Culpa.
567 is a white paste and is infused with teflon particles and stays semi solid. I had to go out to the garage and check the tube for the number.
This is the one with all the "magical" properties I referred to in previous posts.
Best thread sealer I have ever used.

Going back to previous posts and trying to correct the product numbers.

That explains a few things. I couldn't understand how anyone could get an instant seal with 545. Your 567 looks like a very expensive alternative to Rectorseal #5. Hopefully it's worth the extra money.