Resveratrol in the fruits of bilberry (Vaccinium myrtillus L.), the lowbush "wild" blueberry (Vaccinium angustifolium Aiton), the rabbiteye blueberry (Vaccinium ashei Reade), and the highbush blueberry (Vaccinium corymbosum L.) were measured using a new assay based on high-performance liquid chromatography-tandem mass spectrometry (LC-MS/MS). The LC-MS/MS assay provided lower limits of detection than previous methods for resveratrol measurement, 90 fmol of trans-resveratrol injected on-column, and a linear standard curve spanning >3 orders of magnitude. The recoveries of resveratrol from blueberries spiked with 1.8, 3.6, or 36 ng/g were 91.5 +/- 4.5, 95.6 +/- 6.5, and 88.0 +/- 3.6%, respectively. trans-Resveratrol but not cis-resveratrol was detected in both blueberry and bilberry samples. The highest levels of trans-resvertatrol in these specimens were 140.0 +/- 29.9 pmol/g in highbush blueberries from Michigan and 71.0 +/- 15.0 pmol/g in bilberries from Poland. However, considerable regional variation was observed; highbush blueberries from British Columbia contained no detectable resveratrol. Because blueberries and bilberries are often consumed after cooking, the effect of baking on resveratrol content was investigated. After 18 min of heating at 190 degrees C, between 17 and 46% of the resveratrol had degraded in the various Vaccinium species. Therefore, the resveratrol content of baked or heat-processed blueberries or bilberries should be expected to be lower than in the raw fruit. Although blueberries and bilberries were found to contain resveratrol, the level of this chemoprotective compound in these fruits was <10% that reported for grapes. Furthermore, cooking or heat processing of these berries will contribute to the degradation of resveratrol.