It has been agreed between National Laboratories of countries of Suramet (CESMEC- Chile, INTI-Argentina, INMETRO-Brasil, LATU-Uruguay) to perfom a Suplementary comparison in 2000. NIST agreed to participate as an invited laboratory. LATU accepted the role as pilot laboratory. The weights used for the comparison were chosen among those available for comparison in LATU and INTI, that could cover the range of weights mostly used in labs involved. They were of four nominal values: 1 g, 10 g, 100 g and 1000 g. The calibrations were carried out from 02/2001 to 09/2002. All laboratories have reported their values to pilot lab by 12/2002. Data of volume of weights and their uncertainty were available to all participants in the “Measurement Protocol”, and were not measured by the participants. Magnetic susceptibility of weights were in accordance to E2 class. The stability of the standards was checked over a period of one year . The reference value was taken as the weighed mean of all labs. Consistency check was performed, assumed normal distribution. In the cases that a signifficant variation was detected, the reference value was corrected taking into account the estimated drift. Normalized errors were calculated for all participants taking into account E1 and E2 uncertainties according to OIML R 111-04. The results are O.K for 1000 g , 100 g and 10 g weights considering E1 uncertainties and for 1 g weight considering E2 uncertainty. Equivalence between laboratories was evaluated. Bilateral coefficients (bilateral normalizad errors) was estimated taking into account the maximum between declared uncertainty and E1 uncertainties for 1000 g and 100 g weights and E2 uncertainties for 10 g and 1 g weights. There is a good equivalence between participants of comparison.