What a Cold Pan Teaches

I have set a piece of bacon down on a stainless pan that is unmistakably cold to the touch — the metal is at room temperature, the burner has not been lit, the surface has the small dullness that a cold pan has under the light — and I am about to turn the flame on under it. This is not how I was taught to start bacon. I was taught to preheat the pan, get it past the temperature where water would skitter on it, then lay the bacon down and listen for the immediate sizzle. I did it that way for years. Then a cook I worked next to for a long time pointed out, without making a fuss of it, that bacon starting in a cold pan rendered more cleanly, browned more evenly, and did not curl. I tried it, and they were right, and I have been starting bacon in a cold pan ever since. What the cold pan taught me, beyond the bacon itself, was that there are foods whose first lesson is given only at the start of heating, and that the cook who starts everything hot misses that first lesson and has to guess about the rest of the cook from a more limited record.

I do not pretend to have the physics of cold-start cooking pinned in any thorough sense. My best guess after years of doing it on different things is that the cold pan allows the food to undergo a long slow phase before any browning chemistry begins. Bacon, starting cold, has time to release its subcutaneous fat into the pan before the protein on its surface has hit the temperature where it would seize and contract. The fat that comes out becomes the cooking medium that the rest of the cook happens in, which means the bacon is essentially being shallow-fried in its own rendered fat from the middle of the cook onwards, rather than being dry-roasted on hot metal until the fat starts to drip late. A piece of bacon laid on a preheated pan, by contrast, hits the heat and contracts immediately; the protein bands seize and curl the strip into a wavy shape, the fat starts to render but is now being rendered out of a curled and tightening matrix that traps some of it, and the cook ends up with bacon that is browned in patches and lighter where the curl lifted off the surface. The cold-start bacon stays flat. The fat comes out the way the bacon wants it to come out, and the brown arrives later, more evenly, and without the cook needing to flip the strip every thirty seconds to keep up with the curl. I would not swear to any of the temperatures and would not want to. What I would say is that the result, on a hundred pans of bacon, is reliable, and the cook who has done it both ways knows which one they prefer.

The pivot, then, is the patience the cold start asks for. I have come to think of cold-start cooking as a lesson in not rushing the first minute. The cook puts the bacon in the cold pan, turns the flame to medium-low, and walks away for two or three minutes. Nothing visible happens at first. The bacon sits. The pan warms slowly. Then a faint translucent slick appears under the bacon — the first fat coming out of the strip — and the surface of the strip begins to look slightly damp where it sits in the rendered fat. A minute later there is a small steady hiss as the wet edges of the bacon begin to do the things that bacon does, but the bacon has not curled. The fat is now an active medium under it. The cook is not flipping yet. Another minute. The hiss steadies. The bacon begins to take colour at the edges where it meets the fat, but the centre is still pale because the centre is being protected by the strip's own moisture, which is escaping into the air rather than browning on the metal. The cook flips when the underside has the colour the cook wants. The pan, after five or six minutes, has done what a preheated pan would have done in two — but the bacon is flat, the colour is even, the fat is clean, and the cook has had time to do other things while the pan was warming, instead of standing over a curling strip with a fork.

This matters because the most common failure of bacon at home — and the same principle holds for a handful of other foods — is the one where the cook tries to cook the food at the speed they were taught to cook it at, even when the food is asking to be cooked slower. The cook preheats the pan. The bacon goes on. The bacon curls. The cook fights the curl with a fork or a press. The bacon browns in patches, with some pieces nearly burnt and others still soft, and the cook works hard for an indifferent result. The trouble is not that the cook is doing the wrong thing in any obvious sense; the trouble is that the cook is doing the standard thing for a piece of food that benefits from the non-standard thing. The cold start is not the answer to every cooking question. But it is the answer to a particular question — how do I get a fatty piece of food to release its fat cleanly and to brown evenly without curling — and a cook who never tries it never finds out that the question even has a different answer than the one they were given. Bacon was the first one I learned. Sausages are another, especially small breakfast sausages that split if they hit a hot pan. Duck breast, skin-side down, is another — the cold start lets the subcutaneous fat under the skin render before the skin tightens and curls, and the result is crisper skin and less flabby fat than any hot start I have tried. Each of these foods is teaching the cook a related lesson, and the cook who learns the lesson on one of them starts seeing the same shape in the others.

I should name the caveats. Not every food cold-starts well. A piece of fish does not want a cold start; it needs an immediate sear to set the protein on the contact surface and keep the flesh from sticking to the pan as the cook tries to flip it, and the cold start gives the fish too much time to bond to the metal before any browning chemistry begins. A vegetable that is being sautéed for flavour rather than rendered for fat does not want a cold start either; the vegetable wants the high heat that drives off the water on its surface fast and starts the Maillard chemistry quickly. A steak does not want a cold start — the cook needs the contact between the protein and very hot metal to be immediate, and the cold start would leach water into the pan before any crust formed. The cold start is, in my experience, a tool for foods with subcutaneous fat that the cook wants rendered, and it is a poor tool for foods without that fat. None of these break the principle, which is that the cold pan teaches a particular lesson, and the cook who has the lesson is better at choosing when to apply it and when to set it aside.

The bacon is done. It is flat, evenly browned, sitting in a small pool of clean rendered fat that I will pour off into a jar in the fridge for the next time something wants frying. The pan is hot now, finally, ten minutes after I turned the burner on, but it got hot in the company of food rather than alone. I lift the strips off onto a rack, wipe the pan with a piece of paper, and the work is finished. The cold pan was the teacher. The strips are the proof. The next time I do this it will go the same way it just did, and the time after that the same again, and the cold start, which I came to slowly, will keep paying me back for the rest of my cooking life.