How to organize a wine cellar: racks, bins, and a map you'll actually keep updated

My spreadsheet said Rack B, second shelf, and it was telling the truth. What it didn’t say was that the bottle was third from the back behind a case of Xynisteri I’d bought since, so I stood there at eight in the evening unstacking half a shelf to reach one bottle of Maratheftiko while dinner went cold.

Nothing was wrong with the data. Rack B was correct. Second shelf was correct. The location was accurate and it still cost me twenty minutes and a cold dinner, because a shelf is not a place — it is a row of places, several deep, and the record described the row.

That is the whole problem this guide is about. Cellars are not catalogued badly because people are careless. They are catalogued at a resolution that works for the first fifty bottles and quietly stops working after that, and the failure only shows up when you are in a hurry.

Draw your cellar on paper and you will naturally draw a grid: shelves down, slots across. Two dimensions, which is what a shelf looks like from the front.

But a rack is three. Bottles sit behind other bottles, and in a double-depth rack half your collection is invisible from the doorway. A two-dimensional record of a three-dimensional cellar is not slightly imprecise — it is missing the coordinate that determines how much work retrieval takes. “Shelf 3, slot 7” and “shelf 3, slot 7, third back” are the difference between reaching in and unloading.

Some racks add a fourth: bottles stacked on top of one another within a bin. If yours does, that is another coordinate, and the same rule applies — record it or accept that you will rediscover it by hand.

The temptation when setting up any system is to describe the cellar you wish you had: uniform racks, equal shelves, everything on a grid. Real cellars are not like that, and the compromises are exactly where bottles go missing.

• Fridges have unequal shelves. The top and bottom of a wine fridge usually hold fewer bottles than the middle. If your system insists every shelf holds twelve, you will either lose the extras or invent phantom slots.
• Bins are not rows. A diamond bin holds a heap, not a sequence. Trying to number the positions inside one is a fiction; recording it as a bin with a capacity is honest.
• Rooms are separate from racks. A cellar, a kitchen fridge and a case at your parents’ house are three places, and a bottle “in the cellar” means nothing if there are three of them.

The cellar map exists for this: places, units, shelves, bins and diamonds, with capacity calculated from the geometry rather than typed in and hoped for. A cellar cannot claim more room than it has.

Before deciding what goes where, find out what the room actually does. Almost every storage space has a gradient — warm air rises, so the top of a rack can run two or three degrees above the bottom, and a wall shared with a kitchen or a boiler is warmer than the one against soil. Near a door or a window there is also a daily swing.

That gradient is a resource once you know it. Put the bottles you intend to keep longest in the coolest, steadiest part of the room, and the ones you will open within a year wherever is convenient — near the door, at the warm top, in the fridge. You are not trying to make the room uniform. You are trying to spend the good storage on the bottles that need it.

A cheap min/max thermometer left in the room for a fortnight tells you more than any amount of theory. What matters is not the average but the swing: a steady 16°C is kinder than a space that oscillates between 11°C and 17°C every day, because it is the movement — wine expanding and contracting past the cork — that ages a bottle badly rather than warmth alone.

Two other constants, worth stating once. Bottles under cork lie down, so the cork stays wet; screwcap and most spirits do not care and can stand up. And light, particularly daylight and fluorescent, damages wine through glass — which is why the palest rosé and Champagne, in the clearest bottles, want the darkest shelf you have.

Here is the test for any organising scheme, and almost nothing passes it: what happens when twelve more bottles arrive?

Organise by region and a case of Bordeaux means shuffling everything after B. Organise by vintage and the same. Organise by producer and you will spend an evening resequencing a rack every time you discover a new estate. These schemes are lovely in a photograph and unmaintainable in a house.

The scheme that survives is position as an address, not a category. Slot 7 is slot 7 regardless of what is in it. A new case goes wherever there is room, the record says where that was, and nothing has to move. You lose the pleasure of a rack sorted by appellation. You gain a cellar that is still accurate in March.

If you want grouping as well, group by drinking horizon rather than by origin: the things you will open this year near the door, the long sleepers at the back. That grouping changes slowly, which is exactly what makes it survivable.

A map only works if the room agrees with it. Shelf 3 has to say Shelf 3 in the cellar, or you are counting shelves from the floor every time and getting it wrong in dim light.

This is unglamorous and it is the step that makes everything else hold. Number the units, number the shelves, and put the numbers where you can read them with a bottle in one hand. CellarPals prints bin labels straight to a DYMO for this reason — no driver, no desktop software — but a marker pen on masking tape works. What matters is that the physical cellar carries the same names as the record.

Per-bottle labels are a further step and worth it past a few hundred bottles: a code on the neck tag means you can scan a bottle you are holding and learn what it is and where it belongs, rather than reading a label in bad light.

Setting this up for a cellar that is already full feels like the hard case, and it is the one people put off indefinitely. It is genuinely about an hour if you do it in the right order, and the order is: the room first, the bottles second, and never both at once.

1. Describe the furniture with the bottles still in it. Walk the room and record the units — this fridge, that rack, the shelf in the hall — with their shelves and rough capacities. Touch nothing. You are drawing the map, not the territory, and this is fifteen minutes.
2. Number the shelves physically. Tape and a pen, right now, while the layout is fresh. Ten minutes.
3. Then place bottles, one shelf at a time. Take a shelf, record what is on it, move to the next. A shelf is a small enough unit that an interruption costs you nothing — which is the entire reason this works and the “do the whole cellar” approach does not.

If a bottle is already in your record but you cannot see it, leave it unplaced and carry on. Do not stop to solve mysteries; the mysteries mostly solve themselves two shelves later, and the ones that do not were going to cost you the session.

A system that demands a location for every bottle at the moment of entry will be lied to. A case arrives, you are in a hurry, and you type something plausible rather than something true — and a plausible location is worse than none, because you will trust it.

Allow unplaced as a real state. The bottle is in the house, it is in the count, and it does not yet have a home. Then the question “what needs putting away” has an answer, which is the thing that gets the job finished on a Sunday rather than never.

The corollary: a bottle you cannot find should be marked unplaced rather than left claiming a slot it is not in. A record that admits it has lost track of something is more useful than one that confidently points at the wrong shelf.

Adding is easy — a delivery is memorable and you are holding the bottles. Consuming is invisible but at least it is a moment you notice. Moving is the one that quietly rots a cellar record, because reorganising a rack is a physical job with no natural point at which you update anything.

The practical answer is to make relocation cheap enough to do as you go, and to move in batches the record can follow: empty a shelf, refill it, then reconcile once. What does not work is shifting forty bottles over an afternoon and intending to write it up later.

It is also the moment to be suspicious of your own tooling. If the thing you use to place bottles does not show you which slots are already taken, it will happily let you record two bottles in the same hole — and the first you will know is when the shelf disagrees with the screen.

Not tidiness. The return is that what is ready to drink becomes actionable: a list of bottles at their peak is interesting, but a list that tells you which shelf to walk to is something you act on before dinner rather than after.

It also makes the cellar honest about space. Knowing you have eleven free slots in the dining room and none in the office is what stops the next case living in the hall — which, as it happens, is where the next guide starts.

Everything above is how CellarPals works, because it is the cellar I keep. Fourteen days free, then $12.99 a month or $119.99 a year.

If you are moving from somewhere else, the CellarTracker comparison is honest about what carries over and what does not.

Drinking windows explained — When to open the bottles you're saving.

All guides