[en] The present intercalibration exercise provides, once again, strong evidence for insufficient data quality for chlorinated hydrocarbons in marine samples. This comment must be considered in perspective. The principle aim of these exercises is not just to obtain narrower confidence intervals for a given parameter moreover it is to guarantee that the data generated by monitoring exercises is of sufficient quality to evaluate contaminant levels, gradients and trends in the environment. All data should be accurate but precision (expressed as confidence limits) may vary according to its application. It is clearly not the same problem to measure DDT for human health protection (legal concentration limits in seafood range from about 1000-5000 ng/g) as it is to monitor environmental trends where values in biota are commonly one to three orders of magnitude lower. Even the 8 laboratories achieving 'good' data for pp'DDT would not be able to statistically distinguish a 35% increase of the concentration of this parameter from 30-41 ng/g on the basis of the precision observed in the present exercise. Fortunately spatial gradients for DDTs tend to be much larger than this and significant changes could be easily detected by the 'good' labs provided that they use adequate quality control procedures