There was an interview with one of the lead authors of the salmon study on the Canadian Broadcasting Corp.'s science program on Saturday, of which I heard the majority. His take was quite a bit more nuanced (when he had a few minutes to explain himself) than was the impression I got from the article in the newspaper about his study. (And, while I agree with Katherine about scientists over-plugging their own work,, I don't think it happened here: all they did was to buy salmon from all around the world, and test it for organochlorides. These authors didn't establish the link between organochlorides and cancer; that's from other, more well-established work.) For example, he gave a reasonable hypothesis about why farmed salmon from Chile was least polluted, which is that S. America has itself been industrialized far less long than N. America (which had moderate levels of PCBs in the salmon), which has itself been less long industrialized than W. Europe (which was the source of the most polluted salmon). His take was also that farmed salmon could be less contaminated if it were fed a feed that was, itself, less polluted. The working hypothesis is that farmed salmon is less contaminated than the wild because the wild's diet is of fish like shrimp that are themselves not very contaminated, while the farmed's diet is protein and fat from anchovies and other finfish that are contaminated. He also argued that people should eat more salmon, just the wild stuff, and more omega-3-rich fish in general. I mention this in this thread because it influences how I respond to the article. I'm a scientist myself (and I've been interviewed, and mis-represented a couple of times by mass media). As such, I like to get my news as raw as possible when it comes to food science questions. In this case, my response is likely to be small: I really prefer wild salmon's flavor anyhow, and I've eaten a small amount of fish from Lake Michigan which is quite high in organochlorides as well. But I think the "mixed messages" and such entirely come from a mass media that *wants* there to be mixed messages. This study, for example, doesn't give mixed messages; it says that farmed salmon could contribute to higher cancer risk. The mixing comes when the media conjoins this research with previous work on fish / heart disease or fish/hypertension.