The physical evidence is the strongest part of the case, because there is so little of it. No chastity belt has ever been recovered from a medieval archaeological context. What exists is a modest number of objects in museum collections, acquired in the nineteenth century, generally without provenance, and dated by later examination to the early modern period at the earliest and in most cases to the nineteenth century itself.
The pattern behind them is well documented. The nineteenth century had a large private market in curiosities of the barbarous past, exhibitions that catered to it, and metalworkers willing to supply what the market wanted. Several famous specimens have been re-examined and re-dated by the institutions holding them, and museums including major national collections have withdrawn theirs from display or relabelled them as later fabrications. A belt long associated with a fifteenth century Italian lord is among those now catalogued as a modern piece.
There is a second-order point that matters for anyone using museum collections as evidence. These objects were catalogued in good faith, displayed for decades, reproduced in books, and became the illustrations that proved the practice existed. The circularity is complete: the myth produced the objects, and the objects were then cited as proof of the myth. Re-dating them was slow, unglamorous work, and it is why the entry exists in this atlas as a debunked claim rather than as a historical practice.