With a view to developing a new dating method based on extinct 6.8 m.y. Pd107, we isolated silver from 18 samples of the iron meteorites Bristol, Canyon Diablo, Grant, Odessa, Piñon, Sandia Mountains, and Toluca. The silver concentration (11 to 197 ppb) was determined on an aliquot by isotope dilution, and also by neutron activation analysis of separate samples. The isotopic composition was measured by surface ionization on the Argonne 100‐inch‐radius mass spectrometer. No enrichment of Ag107 was detected in any of these samples: our δAg107 values for Canyon Diablo and Toluca were 1.2±2.3 and 3.5±3.7 per mil, compared with Murthy's values of 40±4 and 22±4 per mil. The grand average of all our analyses was 0±1.5 per mil. That this discrepancy was not due to contamination is shown by our low blanks (<5 per cent), by the agreement of the neutron activation and isotope dilution results, and by other evidence. Sampling error is also not a likely explanation, since its probability is <1.6 × 10−3 under the most favorable assumptions. Possibly the apparent anomalies found by Murthy are in some way related to the anomalous value he obtained for terrestrial silver (Ag107/Ag109 = 1.064±0.002, compared with 1.082±0.002, the average ratio found by four other investigators).