The nickel contamination of the metallic patch test substance from 3 commercial sources was estimated by atomic absorption spectrophotometry. Contamination was generally greatest in the cobalt patch test substances. To determine whether the levels of contamination encountered were sufficient to produce a false positive patch test, 20 nickel‐sensitive patients were patch tested to various concentrations of free elemental nickel in aqueous solution. The level of nickel contamination was less than that required to produce a false positive patch test in our patients, but was at a level which could produce false positive reactions in subjects highly sensitive to nicke