Use roughly 5 beans per 8 fl oz (240ml) of 80–100 proof alcohol — that approximates the FDA's single-fold extract strength. Buy Grade B beans, not Grade A: they're drier, cheaper, and carry more vanillin per gram, and nobody's looking at the pod anyway. Split the beans, submerge, seal, and wait at least 8–12 weeks — but six to twelve months is where it gets genuinely good.
Real vanilla extract is one part farming and nine parts patience. There is no cooking, no equipment beyond a jar, and no skill involved — just the right ratio, the right beans, and enough time to let alcohol pull out what the pod spent six months building. Which means the only three ways to get this wrong are using the wrong grade, using too little bean, and being impatient. Here's how to avoid all three, plus the mistakes that quietly ruin a batch after you've already done the hard part right.
Grade B, split lengthwise
~240ml
ready in 8–12 weeks
Scale it linearly. Want double-strength ("two-fold") extract? Double the beans, not the time.
Buy Grade B — This Is the Whole Trick
Most people making extract at home reach for Grade A gourmet beans, because Grade A sounds like the good ones. For this purpose, it's a straightforward waste of money.
Grade A beans are longer, plumper, and hold 30–35% moisture. You're buying that water by weight. Grade B "extract grade" beans are shorter, drier (often under 25% moisture), and cost considerably less — which means a given weight of Grade B contains more actual bean, and more vanillin, than the same weight of Grade A. They're called extract grade for a reason, and it isn't that they're the leftovers.
And the entire argument for Grade A — that it's beautiful, supple, and perfect for scraping seeds onto a plate — is irrelevant here. Your pod is going to sit in a dark jar for a year and then be thrown away. Nobody will ever see it. Buying Grade A for extract is paying a premium for a photograph that will never be taken.
Choosing Your Alcohol
| Alcohol | What it does | Best for |
|---|---|---|
| Vodka (80 proof) | Neutral. Lets the vanilla be the only voice in the room. | The default. Pure, clean, all-purpose extract. |
| Bourbon | Adds caramel, oak, and warmth | Baking — cookies, cakes, banana bread |
| Rum | Adds sweetness and a tropical roundness | Tropical bakes, custards, rum-adjacent desserts |
| Brandy | Adds fruit and depth | Rich pastry, fruit desserts |
Stay in the 80–100 proof (40–50% ABV) range. Below 35% ABV you fall short of the alcohol concentration needed to extract efficiently — and below the level that legally defines pure vanilla extract in the US. Above 100 proof isn't dangerous, just unnecessary; higher proof extracts a little faster but doesn't produce a better result.
The Chemistry: Why the Ratio Isn't a Suggestion
It helps to understand what's actually happening in the jar, because it explains why shortcuts fail. Fresh-cured vanilla pods don't contain much free vanillin — most of it is locked up as a glycoside, bonded to a sugar molecule, and inert until it's released. Curing partially frees it through enzymatic and oxidative reactions over months on the farm. What's left in the pod when you buy it is a dense matrix of vanillin, over 250 secondary aromatic compounds, resins, and sugars, all bound into the pod's cell structure.
Ethanol is unusually good at pulling this apart. It's polar enough to dissolve vanillin and its water-soluble cousins, but non-polar enough to also grab the resinous, fattier aromatic compounds that give real vanilla its depth — the ones water alone can't touch and that a single synthetic molecule can't replicate. That's the entire chemical case for alcohol over any other solvent, and it's also why proof matters: too little water-alcohol balance and extraction stalls; too little alcohol altogether and you're basically steeping beans in flavored water.
Time is doing real diffusion work, not just "steeping." Aromatic compounds have to migrate out of the pod's cellular structure and into the surrounding alcohol, and that migration is slow — which is exactly why heat is never the answer here. Heat would speed up the wrong things: it drives off the most volatile, most delicate top-note compounds before they've had a chance to integrate, leaving a flatter, more one-dimensional result. Patience isn't a folk tradition. It's the correct engineering solution to a slow diffusion problem.
The Method
Split the beans lengthwise
Slice each bean down its length with a sharp knife, leaving about half an inch intact at one end so the pod stays in one piece rather than falling apart into your jar. This exposes the interior surface and the seeds to the alcohol — which is the entire point, since surface area drives extraction speed.
Submerge completely
Put the split beans in a glass jar or bottle and cover them entirely with alcohol. Any pod sticking above the liquid line will dry out and can grow mould, which will spoil the whole batch. If a bean is too long for the jar, cut it — nobody is judging the aesthetics of your extract.
Seal and store in the dark
A tight seal, a cool dark cupboard, and no direct light. Light degrades the compounds you're trying to extract, which is why commercial extract comes in brown bottles.
Shake weekly, and wait
Give the jar a shake once a week. Usable extract at 8–12 weeks. Good extract at 6 months. Genuinely excellent extract at a year. The colour deepens from pale straw to dark amber as it develops, so you can watch it happen — and this is the only step where doing nothing is the correct technique.
Common Mistakes That Ruin a Batch
Most failed batches aren't failures of technique — they're one of a handful of predictable errors, each with an obvious fix once you know what caused it.
Usually harmless — it's fine particulate from the seeds or resins suspended in the alcohol, especially common with lower-proof spirits like some rums. It doesn't affect flavour or safety.
Fix: Strain through a fine mesh or coffee filter before use if appearance matters. If it's using cheap flavored liqueurs rather than plain spirit, switch to a neutral base.
Almost always a ratio problem, not a patience problem. Too few beans per volume of alcohol will never fully develop, no matter how long you wait — there simply isn't enough vanillin present to reach a meaningful concentration.
Fix: Add more beans to the existing jar rather than starting over. It's not too late; the alcohol will keep extracting from the new pods.
This happens when part of a bean sits above the liquid line, exposed to air and moisture, rather than fully submerged.
Fix: Discard that bean and top up the alcohol to fully cover the rest. Push beans down with a clean utensil or a small glass weight if they float.
Common in the first 8–12 weeks — it simply hasn't had time to balance yet, especially with higher-proof spirits, which taste sharper early on.
Fix: Wait longer. The alcohol "bite" softens and integrates as extraction continues; six months makes a dramatic difference over three.
Homemade vs. Store-Bought: Does the Math Actually Work?
The honest answer is: it depends what you're comparing against. Against imitation vanillin, homemade real extract isn't competing on price — it's a different product entirely. Against premium store-bought pure extract, the comparison gets more interesting.
The gap widens further once you factor in the perpetual-jar approach below: a homemade batch that's continually topped up rather than emptied and discarded amortizes its bean cost over years, not months. The only real cost is the up-front wait — which is why most people who make the switch keep at least two jars going on staggered timelines, so one is always ready while another is developing.
Homemade extract improves indefinitely and never really expires — the alcohol makes it self-preserving. So don't finish it. When the jar is about a third empty, top it up with fresh alcohol and drop in a new bean or two. You end up with a perpetual extract that gets deeper and more complex for years, like a sourdough starter that gets you slightly drunk. Add your scraped-out pods from other recipes to it as well; they still hold plenty of flavour.
Scaling It Up: Batches, Gifts, and Bulk
Once the ratio clicks, there's no reason to stop at a single jar. A quart-sized batch (roughly four times the base ratio — 20 beans to 32 fl oz of alcohol) costs proportionally the same per ounce but means you're only managing one steeping timeline instead of several, which is the version most serious home bakers eventually land on.
It's also one of the more genuinely appreciated homemade gifts, precisely because almost nobody actually does it — a small bottle with a handwritten date, given at the eight-week mark with a note that it'll keep improving, reads as thoughtful rather than as a fallback gift. If you're gifting, bottle from the main jar rather than starting a separate small batch just for giving; smaller volumes extract at the same rate but leave you with less margin if a pod ends up sitting slightly above the liquid line.
Frequently Asked Questions
How many vanilla beans do I need to make extract?
Roughly 5 beans per 8 fl oz (240ml) of alcohol, which approximates the FDA's single-fold pure vanilla extract strength. Scale linearly for larger batches, and double the beans — not the steeping time — if you want a stronger two-fold extract.
What grade of vanilla bean is best for making extract?
Grade B, or "extract grade." It's drier than Grade A (often under 25% moisture versus 30–35%), which means more actual bean and more vanillin per gram — and it costs less. Grade A's advantages are appearance and suppleness, which are irrelevant when the pod will sit in a dark jar and then be discarded.
What alcohol should I use for vanilla extract?
80–100 proof (40–50% ABV). Vodka is the neutral default and lets the vanilla speak alone. Bourbon adds caramel and oak, rum adds sweetness, brandy adds fruit and depth. Below 35% ABV the alcohol won't extract efficiently, and that's also the legal minimum for pure vanilla extract in the US.
How long does homemade vanilla extract take?
It's usable at 8–12 weeks, good at six months, and genuinely excellent at a year. The colour deepens from pale straw to dark amber as it develops. Shake the jar weekly, keep it sealed and in the dark, and otherwise leave it alone — patience is the only technique required.
Does homemade vanilla extract expire?
No — the alcohol makes it self-preserving, and it improves with age. Rather than finishing a jar, top it up with fresh alcohol and a new bean when it's about a third empty. You end up with a perpetual extract that deepens for years, and you can add scraped-out pods from other recipes to it as you go.
Why is my homemade vanilla extract cloudy?
Usually just fine particulate from the seeds and resins suspended in the alcohol, which is harmless and doesn't affect flavour. Strain it through a fine mesh or coffee filter if you want a clearer appearance; it makes no difference to how it tastes.
Is homemade vanilla extract cheaper than store-bought?
Usually, yes, against premium pure extract — roughly $1–2 per fl oz at home with fairly-priced Grade B beans versus $4–8+ retail for boutique pure extract. The gap widens further if you maintain a perpetual jar, topping it up indefinitely rather than buying a new bottle each time it runs out.