A stretched bolt should not be reused
What yielding does to a fastener, and why the second use is unpredictable.
A bolt that has been stretched past its elastic range should not be reused because it has been permanently deformed rather than merely tightened, and a permanently deformed bolt no longer has a reliable, known relationship between the torque applied to it and the clamping force that torque actually produces, which is precisely the relationship every subsequent tightening depends on being trustworthy.
An overstretched hair tie
Stretching a hair tie firmly past its comfortable range, just once, often leaves it looking much the same afterwards, but it rarely recovers its original snap, and stretching it again to a point that felt safe before can make it give way without warning. Some of its reserve was used up the first time, and nothing about its appearance records how much.
A bolt tightened within its intended range stretches elastically and returns to its original length when the load comes off, and every later tightening within that range behaves the same way, a known torque giving a known stretch and clamping force. That repeatability is what makes torque a trustworthy way to control preload. Push the stretch past the elastic limit, by over-torquing at assembly or by a large load in service, and the metal yields, lengthening permanently. The bolt may look entirely normal, with the same thread, length and head marking, yet tightening it again risks stretching it further into whatever reduced capacity remains, and there is no reliable way to tell from outside how much that is. Its threads may also have lengthened slightly along the stretched section, so the nut no longer runs as freely, and part of the next tightening torque is lost to that extra friction before any of it produces clamp.
Bolts made to be used once
Some fasteners are designed from the start to be tightened into their yield range. They are brought to a modest torque and then turned a specified angle further, a method that gives a more precise, repeatable clamping force than torque alone can, because once the metal has begun to yield its pull barely changes with a little more stretch. Many engine cylinder head bolts are made this way, and their instructions state plainly that they must not be reused, however sound they look after removal.
The same rule applies to any bolt known or suspected to have been over-torqued: replace it whenever the joint is undone and remade. A new fastener costs very little next to a joint quietly holding less clamping force than its torque figure suggests. A maintenance record that notes a joint was "retightened" without saying whether the bolt was renewed can mislead, since on paper a fresh bolt and a yielded one tightened to the same torque look identical.
Lightly loaded bolts can be reused
A bolt that only ever works well inside its elastic range, never approaching yield across its service life, can be reused repeatedly without losing reliability, and treating every fastener on a design as single-use regardless of how lightly it is loaded adds cost and waste with no safety benefit. What decides it is whether the bolt's application and tightening method ever bring it close enough to yield that full elastic recovery can no longer be assumed. A hair tie that has only ever been looped gently round a ponytail will go on working for years, and it is the one that was once hauled over something far too big that deserves suspicion, even when the two look identical in the drawer.