We’ve heard the claims: A “break­through in herbaceutical science” (Avon Anew Alternative). “Regenerate damaged skin” (Olay Regenerist). Have your wrinkles “virtually slept away” (Lancôme Rénergie). Americans spend more than a billion dollars a year on those wrinkle potions and others.

In our first test of wrinkle creams (performed in partnership with--who else?--the French), we found that after 12 weeks our top-rated products did smooth out some fine lines and wrinkles. But even the best performers reduced the average depth of wrinkles by less than 10 percent, a magnitude of change that was, alas, barely visible to the naked eye.

Moreover, the luxury-priced skin-care offerings didn’t work any better than the drugstore brands. La Prairie Cellular, the most costly product we tested ($335 for an ounce of day cream and 1.7 ounces of night cream) was among the least effective. Our top-rated product line, available at a drugstore near you, is Olay Regenerist--as it happens, the nation’s best-selling brand of wrinkle cream--which cost about $19 apiece for the “enhancing lotion,” “perfecting cream,”and “regenerating serum” combination that the company recommends.

We also found that:

Our testing lasted long enough for changes to show up if they were ever going to, our consultants said.

We chose a sample of top-selling mass-market lines (Olay, L’Oréal, RoC) and some costlier products (La Prairie, StriVectin). We bought them in retail stores, and covered the labels on all bottles, tubes and jars so that the women wouldn’t know which brand they were using.

We chose women between the ages of 30 and 70 with light skin, the type most susceptible to developing the fine lines and wrinkles that most creams claim to address. Each cream was tested by 17 to 23 women. They were recruited by a European laboratory specializing in cosmetic evaluation.

Each woman used a test product on one side of her face and the lab’s standard moisturizer on the other for comparison (she didn’t know which was which). The moisturizer had a sunscreen but no anti-aging ingredients. We asked users to apply the creams as they normally would, making sure to rub them into the crow’s-feet area, at the outer corner of their eyes. Most product lines consisted of a day cream and a night cream, and one, Olay Regenerist, also included a serum. StriVectin-SD came in only one formulation, which the women used twice a day.

We examined their skin at the beginning of the 12-week test, four weeks into it, and at the end, using various measures:

Both the photo assessments and instrument measurements found, on average, only slight improvements. But every product performed better than that for at least some test subjects and failed completely for others. Given what is known about skin aging, that’s not surprising. “Skin is not Saran Wrap,” says Amy Newburger, an associate attending physician in clinical dermatology at St. Luke’s-Roosevelt Hospital Center in New York City. “It’s alive and metabolically active. Each person’s skin has its own unique requirements and responses.”

We found no relationship between the types of active ingredients in the products and their overall performance. Our consultants say it’s possible that the supposedly inactive ingredients in products can affect the performance and absorption of the active ingredients. Nor did the presence or absence of an SPF rating have any effect; products with and without it appear throughout the Ratings. All of the products were effective moisturizers.

There was no correlation at all between price and effectiveness.

Very few women reported that the products irritated their skin, which is to be expected of over-the-counter compounds designed for safe use by the general population. The exception was RoC Retin-Ox+, which contains retinol as its active ingredient. Doctors who work with retinoids, the stronger prescription version of retinol, routinely counsel their patients to expect initial redness, flakiness, and irritation, and some women who tested the over-the-counter product reported similar symptoms.

When we combined the overall results of all our tests, the products ended up grouped into three categories we’ve termed “slightly more effective,” “average performance,” and “slightly less effective” (see our Ratings).