This report concerns a series of comparisons of neutron activation analysis (naa) results obtained at Hebrew University, Jerusalem and at Lawrence Berkeley Laboratory. The goals of the study were (1) to identify elements which seem to be determined accurately based on high interlaboratory agreement and (2) to initiate a search for what is wrong in cases where there was disagreement. Six comparisons were run, 3 ancient potteries and 3 rhyolytic tuffs. Possible sources of error in naa are first discussed. These include standards, counting error, procedural uncertainties, spectral analysis errors, peak shape errors, and absorption corrections. Interferences are particularly important sources of errors in the spectral analysis, and a number of types of interference are discussed. The procedures used at each laboratory are described briefly. Results are presented as a series of tables, one for each of 23 elements, comparing values obtained on each specimen in the two different laboratories. Each table is accompanied by a brief discussion of the data and sources of error for that particular element. It is concluded that, for elements determined with a precision of less than 1 percent, interlaboratory agreement of results should be good to 2 percent.