The Helsinki City Centre of the Environment tested two methods, dimethylglyoxime (DMG) and atomic absorption spectrometry (AAS), for detecting nickel release in piercing earrings. The DMG screening test was performed in two slightly different ways, with and without ethanol and heat prehandling. All 30 pairs of earrings tested, totalling 66 objects, were negative. However, according to our AAS test, 25 of the 66 objects (38%) released greater than or equal to 0.05% of nickel, the mean amount being 2.1% and the maximum 12%. When measured by AAS after artificial sweat treatment, 11 objects released more than 0.5 mu g/cm(2) per week of nickel, the mean amount being 3.4 mu g/cm(2) and the range < 0.1-84 mu g/cm(2). After this sweat treatment, nine of the objects (14%) were positive in DMG tests. These findings indicate that the DMG test is unreliable for detecting nickel release from jewelry. Quality control of consumer items should be performed by laboratories that have quantitative analysis methods for such investigations. (C) 1998 Elsevier Science B.V. All rights reserved.