Saffron is expensive enough, and hard enough to inspect, that adulterating it has been worth someone is trouble for about as long as it has been traded. The substitutes are good. Safflower petals, dyed maize silk, shredded beetroot fibre and coloured gelatine threads all look convincing in a photograph and most of them look convincing in the hand.
What follows are the checks that actually separate them, in the order they are worth doing, and then the two tests you will see repeated everywhere that tell you nothing.
1. Look at where the thread changes colour
A saffron stigma is a trumpet: narrow at the cut end, flaring into a serrated funnel at the tip. Look at the flare under good light. Real saffron is not one flat colour — it runs slightly deeper at the mouth of the trumpet and slightly paler at the cut.
Safflower, the most common substitute, gives itself away here. Its petals are flat, straight-edged and uniformly coloured, with no trumpet and no serration. Once you have seen the flare you will not confuse the two again.
2. Put a thread in cold water
This is the test worth doing, and the detail most people get wrong is the temperature. Use cold water, not warm.
Real saffron releases its colour slowly in cold water — you will see a fine golden thread of colour descending from the stigma over several minutes, and after fifteen minutes you will have a clear yellow-gold liquid. Crucially, the thread itself stays red. It does not go pale.
A dyed substitute does the opposite. The colour comes off fast, often within seconds, the water turns red or orange rather than gold, and the thread left behind is noticeably bleached.
3. Rub a thread between your fingers
Take one thread and work it between thumb and forefinger for a few seconds.
Real saffron leaves a faint yellow-orange stain on your skin. Dyed material leaves a red one. This is the same principle as the water test and takes ten seconds, so it is worth doing as a first pass on a market stall.
4. Check the smell, then check it again after five minutes
Saffron smells of hay, honey and something faintly medicinal — it is a complicated, slightly bitter smell, and it is not sweet. A sweet, simple, perfumed smell is a warning sign, because sweetness is what gets added when something needs to smell like saffron rather than being it.
The second sniff matters more than the first. Real saffron keeps developing; an added fragrance fades quickly and leaves the material smelling of nothing.
5. Weigh it
This is the least romantic check and the most reliable. Saffron is extraordinarily light. A gram is a surprisingly large heap — roughly 450 to 500 threads. If a gram fits in a small pinch and feels like it has any heft at all, something in the jar is heavier than saffron: usually moisture, sometimes glycerine, occasionally sand.
Buy from someone who sells by weight and states the thread count, and this becomes their problem to get right rather than yours to detect.
The two tests that prove nothing
The baking soda test. You will read that real saffron turns baking soda solution yellow and fake saffron turns it red. Dyes that pass this test are cheap and widely used. It is not a meaningful check any more.
The "it must be expensive" test. Price filters out nothing. Adulterated saffron is priced to look like the real thing, because a suspiciously cheap jar does not sell. Paying a lot is not evidence of anything except that you paid a lot.
What actually settles it
Every test above is a sanity check, not proof. The thing that genuinely settles the question is documentation: a lot number that ties your tin to a specific harvest, and a laboratory report for that lot. Under ISO 3632 a sample is measured for crocin, which is colour; picrocrocin, which is taste; and safranal, which is aroma. Those three numbers are a fingerprint, and no amount of dyed maize silk produces them.
Ask for the report. A seller who has one will send it without hesitating, and a seller who does not have one will tell you at length why it does not matter.
