Precious metals are not very magnetic. So if it's very magnetic it can't be gold (or silver or platinum). Of course, if it's NOT magnetic it doesn't prove anything.
Precious metals are soft, large thin gold items (especially if made of high-carat gold) bend easily in and out of shape. Also, before testing the item you will have to file the surface with a fine steel file (otherwise you will merely test the surface plating). This is a good test in itself, you will soon see how easy it is to file gold or silver whereas the steel file will 'bounce off' an item made of steel - no need to test any further.
Normally you need a set of test needles (one per each karatage being tested), test stone and a set of acids, one for each karat value. Normally you would have to have 10K, 14K, 18K, 22K, silver and platinum acids. You need to be very careful with acids, as they are harmful chemicals. Use gloves, goggles, and an apron. Also, acid testing should only be completed inside a proper fume hood. We use a ductless fume hood with acid filter.
Gold Testing Procedure with Acid Test
Scratch your piece over the surface of a black gold-test stone, pressing well enough to leave a visible deposit. A line of one to one-half inches long is suitable. For those looking for the most accurate testing available it is important that the user becomes familiar with "comparative testing" which uses standard gold testing needles. For the best in sensitivity, leave a scratch line alongside the scratch line of the metal that you are testing with a gold test needle. Take note of the speed at which the scratches dissolve. If your test scratch dissolves more quickly than your needle scratch, you have a lower karat than the test needle.
Leave a drop of the 10K solution on the scratch made. Should the solution dissolve your scratch on the stone this indicates that the object is less than 10K gold or not even gold at all. Should the solution leave your scratch intact, your object is most likely 10K or greater than 10K.
You can repeat as above using the 14K solution. If the solution dissolves the scratch on the stone you will have less than 14K gold (if the scratch dissolves slowly and leaves rusty color particles, it is likely 12k gold.) If the solution does not affect the scratch, your object is 14K or greater than 14K.
You can repeat as above using 18K solution and 22K solution (when available) to find the exact karat of your object.
Keep in mind that when the solution used dissolves the scratch slowly and leaves rusty color particles it is likely two karats lower than the solution being used. Keep in mind that 24K gold cannot be determined with the acid test and that most fake gold will fail the acid test but it is not a foolproof test.
The sample of gold is weighed very precisely and the amount is recorded. The sample is wrapped in assay lead foil along with a quantity of pure silver. This wrapped ball is placed in the furnace in a cupel (a special kind of disposable crucible). All the non-precious metals are absorbed by the hot cupel. The precious metal forms a button within the cupel. The cupel is removed from the furnace, metal removed from the cupel and then brushed to remove any lingering bits of cupel. It is hammered flat, rolled thin and then heated in a porcelain crucible containing a weak nitric acid solution. The acid removes the silver, which is poured off and the silver recovered from solution. The gold is then rinsed in distilled water to remove any residual acid and then dried. The sample of gold is now at least 99.999% pure. The sample is then weighed again. The original weight of the impure sample is divided into the weight of the now pure sample. The result is the assay.
X-ray Fluorescence Assay
The results are 99.99% accurate when the most advanced machines are used. They have made fire assays obsolete. There is no destruction to the pieces. Test will determine if samples meet government standards for 10k, 14k, 18k 22k and 24k. Platinum and silver testing is also done. The entire metal in the sample is tested simultaneously and a breakdown of the other elements will be disclosed as well. i.e. Cu, Ag, Zn, Ni, etc. XRF is also used to identify harmful contents such as lead and cadmium. The test takes less then 90 seconds.
Cheap Electronic Gold Testing
It is affordable (from less than $100 to about $500), fast, accurate, sensitive readings even when operated by a novice, nondestructive.
Testing liquid or gel is required to operate, but those liquids or gels are not harmful to the operator and are not destructive.
Cheap electronic testers have a reputation for being unreliable. Few professional gold dealers rely on these machines.