Black, no sugar, still breaks the fast
The intuition is reasonable: no sugar, no milk, almost no calories, so surely it cannot matter. It does, and the reason has nothing to do with calories. Coffee is not metabolically inert.
In a crossover trial, healthy volunteers given caffeinated coffee before a meal showed a glucose response 147% to 216% larger than the same people given decaffeinated coffee, with insulin sensitivity reduced by up to 40%. In a separate controlled trial, coffee on its own raised the insulin response compared with water. That is enough to disturb precisely what a fasting glucose is trying to observe.
Which tests it disturbs, and which it does not
Fasting glucose is the one that matters most. It is the test the fast exists for, and it is the one coffee moves.
For the lipid panel the picture has changed, and for a reason unrelated to coffee: European guidance has recommended non-fasting lipid testing as the default since 2016, because the differences after an ordinary meal are small and do not change how the result predicts cardiovascular risk. If your laboratory no longer asks you to fast for cholesterol, the coffee question falls away with the fast.
A blood count, a thyroid panel, a liver panel or an HbA1c do not require fasting at all, so a coffee before them changes nothing you need to worry about.
How long to leave, in practice
Treat coffee like food: it belongs on the wrong side of the fast. If you were asked to fast eight to twelve hours, that includes coffee for the whole period, not only the last hour.
Decaffeinated coffee is not a loophole either. The trials above compared caffeinated with decaffeinated precisely because decaf is the milder option, not a neutral one, and it is still a drink with something in it.
The workable plan: coffee after the draw, not before. Many laboratories are used to people arriving with that plan and nobody will think it odd.
If you already drank one
Say so at the reception desk before the needle goes in, not after. The laboratory has two ordinary answers: rebook you for another morning, or draw the sample anyway and note on the form that the fast was broken. Both are better than a fasting glucose that nobody knows was not fasting.
Nothing about this is a wasted trip in most cases. Many of the lines on a routine prescription do not care about a coffee at all, and the laboratory can often run those and repeat only the two or three that need an empty stomach.
What you should not do is stay quiet and hope. A glucose read as fasting when it was not can push someone towards a diagnosis they do not have, or hide one they do.
Tea, and everything else
Tea contains caffeine too, and herbal infusions contain plant compounds that are not water. Sugar-free soft drinks, chewing gum and flavoured water all fall under the same rule.
The only exception is plain water, which is allowed throughout and actually helps: a hydrated vein is easier to draw from.
If you drank a coffee out of habit and only remembered at reception, say so. Depending on the test the laboratory will rebook you or take the sample with a note on the report. Saying nothing is the only bad option.