Gaithersburg, Md.—Uncle Sam can’t stop you from drinking till you’re stupid, but the feds are taking steps to make sure it’s the alcohol that does it to you—not the lead.

Centuries ago, Roman vintners used lead to sweeten their wine. That practice stopped a few hundred years ago, but until it was banned in 1992, the foil used to wrap corks often contained lead that could find its way to an aficionado’s palate. Even today, wine may contain lead taken up from the soil or deposited on grape skins by exhaust from passing vehicles.

The problem is serious enough that the International Organization for Wine and Vineyards established a maximum recommended level of lead in wine of 200 micrograms per liter, and the World Trade Organization has adopted the recommendation. Accordingly, many countries now test imported wines to verify compliance with the standard. Similar standards exist for a host of other foods and beverages.

"From a metrological point of view," according to the draft of the European Commission scope statement describing the study, "the measurement of the amount of lead in wine is representative of many similar measurements in various food and beverage matrices.

"Laboratories who demonstrate their capability of measuring lead content in the CCQM-P12 wine samples are likely to have the capability, knowledge, and skills to measure the amount of other elements at similar levels in other food and beverage matrices."

Worldwide testing for lead content in a uniform wine sample established that the National Institute for Standards and Technology (NIST) laboratories are capable of measuring lead in extremely small amounts.

Though the institute isn’t going into the wine-testing business, the results establish that commercial laboratories will be able to use NIST expertise to develop their own testing programs for wine or other foods or beverages.

"We like to think of NIST as the nation’s reference laboratory," said Dr. Jack Fassett, who directed the testing.

Fourteen laboratories tested the sample: a mix of 500 0.25-liter bottles of Bordeaux from a single batch of wine. The mean result was 0.1326 nanomoles per gram, with a standard deviation of 0.0041.