America's Test Kitchen LogoCook's Country LogoCook's Illustrated Logo

Red Wine for Cooking

Do distinctive flavor profiles come through once the wine is cooked down?

When a recipe calls for red wine, the tendency is to grab whatever is inexpensive or already open on the counter. But as with any ingredient in the kitchen, the choice of wine can make the difference between a sauce worthy of a four-star restaurant and one that's best poured down the drain. In short, the wrong wine can turn an otherwise good sauce bad. The problem is that sifting through the enormous range of wines available is only slightly less confusing than trying to plough through Ulysses. Choosing a good bottle for the kitchen can seem like, at best, a shot in the dark.

To find out which red wines are good cookers (as opposed to those which are just good drinkers), We set up three tests—a quick tomato sauce, a pan sauce for steak, and a long-cooked beef stew—through which we could test numerous bottles. With the help of local wine expert Sandy Block (who holds the title Master of Wine, an honor shared by just 18 Americans), we organized those red wines available in even a poorly stocked wine shop into four manageable categories based on flavor, body, and style: light/fruity, smooth/mellow, hearty/robust, and nondescript jug wine. Ironically, the only type of wine not represented in these four categories is the "cooking wine" found on most supermarket shelves. In previous tests, the test kitchen has found that these low-alcohol concoctions have little flavor, a high-pitched acidity, and an enormous amount of salt, all of which combine to produce inedible sauces. Omitting this type of wine from the testing, we began our search for the ultimate red wine for cooking.

Putting Wines to the Test

We began by cooking with a representative from each of the four categories: a light/fruity Beaujolais, a smooth/mellow Merlot, a hearty/robust Cabernet Sauvignon, and a jug of "mountain" (sometimes also labeled "hearty") burgundy. Although none of the groups emerged as the winner from this first round of tests, what did emerge were some important attributes of good and not-so-good cooking wines. While the sauces made with Beaujolais could be described as wimpy, this wine did contribute a refreshing fruitiness that mingled well with the other ingredients and complemented their flavors.

The Merlot-based sauces had a somewhat overcooked, jamlike flavor, but they were also well-structured. The hearty Cabernet Sauvignon gave the sauces an astringent, woody bite that bullied all other flavors out of the way. This wine is aged in oak barrels, and its resulting oak flavors did not soften as they cooked but turned bitter and harsh. The jug wine, meanwhile, made sauces that were overly sweet and simple. Based on these findings, we decided to try to find wines that would combine the most appealing qualities of the light/fruity and smooth/mellow wines, shying away from wines with an oaky influence and inexpensive jug wines.

Focusing on this more narrow category—fruity/smooth/mellow—we selected four new bottles of wine: Sangiovese (a medium-bodied wine from Italy), red Zinfandel (from California), Pinot Noir (from Burgundy, France), and Côtes du Rhône (from southern France). The Sangiovese tasted great in the tomato sauce but made an astringent pan sauce and a cardboard-tasting stew. The Zinfandel tasted overcooked and jammy in the tomato sauce and turned the pan sauce bitter. While both the Côtes du Rhône and Pinot Noir turned in impressive results across the board, the Côtes du Rhône was stellar. When compared with the sauces made from Pinot Noir (a wine made from just one type of grape), the Côtes du Rhône (a blend of several grapes) had a fuller, more even-keeled flavor. The different grape varieties in the blend appeared to compensate for each other's shortcomings to produce a potent, well-rounded flavor.

We then tested several more fruity, medium-bodied blends with little oak, including wines from the greater Rhône Valley (in southern France), Languedoc (also in southern France), Australia, and the United States. These blends were not made from the same grape types, but they all made tasty, well-balanced sauces. Our conclusion? Most red wines made from a blend of grapes will produce good sauces—just steer clear of wines aged in oak.

Does Price Matter?

Next came the question of price. Would a $30 blend make a better pan sauce for steak than a $5 blend? To find out, we cooked with fruity, medium-bodied red wines made from a blend of grapes at four price points: $5, $10, $20, and $30. Tasters found that the results produced by a $5 bottle were much different from those produced by bottles in the other price categories. As wine cooks and reduces, it becomes an intensely flavored version of itself, making the wine's defining characteristics unbearably obvious. The sweet, bland $5 wines cooked down to candy-like sauces, while the $10, $20, and $30 bottles were smoother, making sauces with multiple layers of flavor. Although tasters favored wines in the two more expensive price ranges for their slightly more balanced and refined flavors, none thought the difference justified spending an extra $10 or $20. And we found that limiting the price to around $10 still left us with plenty of options when shopping. We had consistently good results with several widely available wines, including E. Guigal Côtes-du-Rhône ($9.95), Rosemount Estate Grenache/Shiraz ($8.95, from Australia), and Bonny Doon Ca’ del Solo ($12.25, from California).

Now We're Cooking

As we cooked our way through multiple bottles of wine, we found that it is not only the type of wine that matters but also the way you cook with it. The right wine can taste all wrong if cooked badly. That's because as wine is heated, delicate flavor compounds known as esters break apart, turning fruity flavors and aromas muddy and sour. The higher the heat, the more rapidly these esters will change from good to horrid.

Transferring this knowledge to cooking, reason suggests that wine would best be treated with low, slow heat. In fact, our testing had demonstrated this point. The beef stew was much more forgiving than the tomato or pan sauce, both of which are typically made by means of a fast and furious reduction over high heat. To further test this proposition, we made two more steak pan sauces, one by rapidly simmering the wine, the other by slowly reducing it, just below a simmer. The results were so radically different that tasters thought the sauces had been made from different wines. The sauce made from the rapidly simmered wine was tart and edgy, while that from the slowly reduced wine was round and smooth. The fast reduction had bruised the wine's esters; the slow reduction allowed their true, fruity flavor to shine through.

As we tested a few more pan sauces and did a little more research, we were introduced to another cooking trick by chef and wine importer Richard Kzirian. He suggested adding small amounts of aromatics to the wine as it reduced to add an extra dimension of flavor and polished texture. Treating the wine almost like a stock, we infused small amounts of shallot, carrot, mushroom, parsley, and bay leaf into the reduction. This made for a pan sauce that was rich and voluptuous, with complex layers of flavor.

Although this reduction is cooked over low, slow heat, we found the process can be speeded up by using a skillet rather than a saucepan. The increased surface area of a wide skillet allows the wine to evaporate more rapidly, even when cooked over low heat. We also found it easiest to remove the syrupy reduction from a nonstick pan, although a regular pan also works. The aromatics are strained out before the reduction is fully complete to keep them from soaking up too much of the increasingly valuable wine as it reduces.

To sum up the test results: A good bottle of cooking wine is likely to be made from a blend of grapes and can be had for about $10. The wine should have good fruit flavor, medium body, and little or no oak flavor. In the kitchen, this wine should be cooked just below a simmer with aromatics and not treated to a hot boil. While this technique is not necessary in recipes that already call for cooking wine slowly with aromatics, as in a stew, it makes all the difference with wine intended for a pan sauce or a quick tomato sauce.

This is a members' feature.