There's being some success right now in dating ferrous metals through the carbon found in the trapped slag inclusions and even through the carbon from the carbides in the steel itself, using Accelerator Mass Spectrometry(AMS). See, for example, the works of R. G. Cresswell and, more recently, of Andrea C. Cook and Jeffrey Wadsworth. The use of AMS allows to work with much smaller C amounts than needed before, so it’s opening up the path for analysing a much wider spectra of samples.
Of course, this assumes that the carbon present in the steel to be analysed comes from organic sources, i.e. charcoal. The good news is that a significant majority of antique metals were smelted and worked using charcoal, coal smelting only starting to be prevalent in Europe at the end of the 18th c. Although this is far from being definitive (you knew I was going to say that, did you?), sometimes it is found that some coal was indeed used at some point in the manufacture of the item, then the resulting date jumps back thousands of years, and as the academic community, so far, seems somewhat unconvinced of the existence of an Hyborean Era, one has to conclude that the most likely reason for such inconsistence is the use of mineral fuel.
There are also other things to be taken into account, like the early use of coal in places like China to smelt and cast iron, etc. Caution is always needed when analysing archaeological samples... And keep in mind that radiocarbon dating has quite a margin of error, normally, in this cases, of some decades, so one can’t pretend excessive precision in the results.
For nonferrous metals... well knowing your patinas is a must for any collector of Cu-based alloys items. The techniques to fake patinas are many and well-developed, so, well, you know the mantra:
caveat emptor. For a collector’s opinion on this regard, I would recommend taking a look to the interventions by Kenneth Blair in
this thread in SFI. Also,
this one may be illustrative.
Beyond "field" experience, and as corrosion is extremely variable depending on the particular conditions every item has been subjected to, a moderately well-equipped archaeological/archaeometallurgical lab has enough resources to tell old from ancient. Sometimes, especially if the patina of the item has been messed with, this is not easy and some invasive techniques (i.e. taking samples of the actual metal) must be used. The degree and morphology of the corrosion is studied, the nature and concentration of elements in the patina is checked, the coherence of the traces left by the manufacturing methods are contrasted... well, the whole nine miles, if done properly. This is, also, somewhat expensive. Some items are not worth enough (in market value) to have them go through such exhaustive testing.
In short, there's no simple answer. There's the means, indeed, but, as always, in many occasions no easy way is available. The best bet, for a change, is experience. Becoming acquainted with the basics of the corrosion processes, with the most usual accelerated corrosion and overall faking techniques and the results they yield, with the details of historical manufacture and the features they leave in the final product... in short, as always, educating oneself.
Hope this helps…