It's something in the composition. I can't remember who, but either Seachem or API mentioned it somewhere. There's a recently closed thread started by Mamajin about the differences in accuracy between different test kits. Another test kit shows 0 ammonia, even when Prime is used. Before I knew about it, I disregarded the .25 ammonia reading, because my fish didn't show any of the symptoms of ammonia, and multiple PetSmart tests came up with 0 ammonia on the strips, and my LFS even backed up that reading using a liquid test kit. Hm. Maybe I should ask what test kit they're using.....