The comment about valve springs (tappet springs) is ALSO addressed below -- same point: if a working spring is kept within the spring's design envelope, cycling a spring doesn't do much damage. Some of the newer designs (sub-compact guns and very hi-cap mags) intentionally push springs beyond that design envelope. The springs do MORE for less long; they become renewable resources. But those are special cases.
The topic of mag (and recoil) springs comes up periodically and the comment, "only cycling" wears out springs is trotted out. Working springs CAN lead to wear, and leaving some mags fully loaded can ALSO lead to even greater wear -- and according to the science of spring design, it's not as simple as it seems. I'm repeating below, what I posted earlier in another discussion on another forum, but parts of this have been posted on this forum, too:
According to the experts here and on another forum (The Firing Line), which has included a number of engineers who are familiar with the subject (at least one of them a metallurgist), most springs -- while they will degrade with use -- won't degrade enough to matter, if properly designed and properly used in the gun's design. They'll typically outlive the gun or the shooter. As noted above, however, many of the new SMALL gun designs, or guns with very high-cap mags, often push the springs' limits and springs and those spring can have shorter lives.
When recoil or mag springs do degrade, due to cycling, it's because they have been compressed, near, or beyond that point of maximum (design) compression. If the spring doesn't compress to that point (called its elastic limit), the spring (even if the mag is fully loaded, or the slide is locked back) won't degrade much with use. A lot of gun springs, when cycled, NEVER get close to the spring's elastic limits.
Working a spring alone isn't necessarily a problem: tappet springs in a car engine are an example -- as they may cycle many millions of times without failures over an engine's life... They're designed with reserve power, and unlike springs in some of the newer gun designs, they're not asked to do more work in less space, with less material than was once the practice.
In earlier discussions here and on The High Road, it was noted that the small recoil spring for the Rohrbaugh R9 -- probably the smallest 9mm semi-auto -- should be replaced every 250 or sound rounds. That gun fires the same round as guns that have recoil springs that last thousands of rounds! What's the difference? That small spring is apparently pushed to it's elastic limits with each shot -- and that spring just can't last as long as other less stressed springs (that are made to fit larger spaces and use more metal). Most spring applications do NOT stress springs that much.
The point of maximum compression is when (and where) spring wear takes place. If the spring is kept loaded, and the spring is near it's design limits at that point -- and, arguably, most mag springs are not be near that limit when fully loaded -- then the spring will degrade a bit more quickly than if the mag is downloaded for cycling or storage. For most full-size, non-hi-cap mags, it's not likely to be a problem. For some hi-cap guns, or for some sub-compacts, it can be an issue.
.