BEWARE!!! Especially those who haven't...

[USMC kalibrater]

been around this monster, we refer to as third party calibration!
Do your homework on companies you opt to work for. 
Ask in the forums or pm friends to find people who have worked for some of these "Accredited" companies. 
I try not to talk bad about companies so Ill part with this
I currently work for a great company, Technical Maintenance Inc.  If you are looking check us out. 
However
I have worked for several other 3rd party labs.  Some were ok, one was a complete nightmare/fraudulent company and could have really left me in a bad situation.  If you pm me I will give my opinion on any company I have worked for, most will be good or ok...that one though...whoa

[Hawaii596]

I've been around that block as well.  Those "bottom feeders" really make it tough for everyone.  There is a particular one whose name I won't mention that I recently informally audited 8 certificates of calibration from them.  They, for example, used a Fluke 5500A to calibrate an HP 3458A.  They used only a microwave signal generator to calibrate an oscilloscope.  They used a Fluke 85 as the only standard to cal a Keithley 263 voltage calibrator.  I was told second hand by a tech co-worker that they did a cal with data on a spectrum analyyzer for $167, and they provided three data points (I don't know what those data points were).  I still have the PDF's of the cals I audited, in my personal files.  Odd thing was., for every one of the cals I audited, that they did a completely fraudulent job on, in their "scope of accreditation" they listed all of the required standards as being available.  This was a multi-lab, multi-national company.  I'm not going to mention their name.  But all except the last one, I know factually first hand.  Regarding the spectrum analyzer that I was told about, the co-worker told me he challenged them on it, and they said they did the cal in accordance with their internal procedure.  That is complete and utter nonsense.

I think ISO17025 has done a dis-service to industry in some senses because it can create a false sense of security.  I'd rather see actual audits of actual calibrations done.  Problem with ISO17025 is that companies like that can do the work in accordance with the ISO.  Then customerssay they don't want to pay the extra for the ISO17025 accredited cals.  So scummy companies still do poor, fraudulent work when it is not an ISO accredited cal.

Enough of a vent for one day.  And just so friends know, it was not the multi-lab, multi-national company that I have a friend or friends in.  It was another company.  Also, I have worked for and dealt with a number of third party companies, some big and some small.  There are some good ones out there.  And even with the good ones, it comes down to how the local lab is operated and the quality of technicians on the bench.  There are some major third party companies I know of personally that operate poorly in one location, and very respecatably in others.

[RFCAL]

Yeah, there are bottom feeders out there and they should stay on the bottom. The problem is, if the cal is not a 17025 cal, then A2LA or any other accrediting body cannot do anything about it. They could do everything correct on an audit and still botch a commercial cal with no penalties.

[cjohnson]

I can vouch for Technical Maintenance as well. It's an extremely well-run and properly resourced company led by people who genuinely care for both their employees and their customers. We're up over 100 employees and 5 branches now, so you're bound to run into us at some point, and when you do, don't run the other way... We're always looking for great people to join us.

[spanishfly25]

Another problem with accreditations is that sometimes a lab get certified for ONLY for one type of calibrations like multimeters but the rest isn't certified and they would show the accreditation certificate to everyone making you believe the whole lab is certified.

[NC-Cals]

You must review the scope and uncertainty closely to make sure their accredited uncertainty meets your needs.  They can "calibrate" an 3458A with a 5500A if they report their uncertainty correctly - even if the uncertainty exceeds the accuracy of the 3458A.  Buyer beware for this little accredited calibration trick.  Also, they are not allowed to use thier ISO accreditation on any calibration outside of their scope.  They also cannot report uncertainties tighter than their published scope.  If they are found to do this they can lose their accredation.  If they do a cal that is partially accredited then they must clearly document what is and is not accredited. 
  The only way this ISO accreditation system is going to work is if violators get reported and shut down.

[griff61]

caveat emptor

Bottom line is know what you are buying when you are getting service.

[sdmetrol]

As an owner of a calibration lab, I recommend that before utilizing a lab, one should look at the structure of the company.  At a minimum, work should be assigned by a senior technical manager, followed by quality approval by a fully qualified metrology professional.  Training records should be readily available with sufficient evidence of technical training for the technician listed on any certificate being audited.

The examples listed by Hawaii, if true,  are a disgrace to our industry and show an obvious lack of control of the processes by the management of that lab.  If the management and/or ownership is not directly responsible for that type of output, it is no excuse as they are certaintly responsible for the focus, guidance, and maintenence of the integrity of their lab.