Perfluoroketones, used as replacement to halons and CFCs, are excluded from the Montreal Protocol because they are considered as nonozone depleting substances. However, their chemical structure makes them possible greenhouse gases if their atmospheric lifetimes are long enough. To assess that possibility, we investigated the photolysis of perfluoro-2-methyl-3-pentanone (PF-2M3P), and perfluoro-3-methyl-2-butanone (PF-3M2B) using outdoor atmospheric simulation chambers. In addition, the photolysis of a non fluorinated pentanone (2-methyl-3-pentanone, 2M3P) was studied. The results showed that photolysis is the dominant loss pathway of PF-2M3P and PF-3M2B in the troposphere whereas 2M3P is lost by both photolysis and gas phase reaction with atmospheric oxidants. The photolysis effective quantum yields of PF-2M3P, PF-3M2B, and 2M3P were estimated and some of the main products identified. The photolysis of PF-2M3P and PF-3M2B was found to have a minor impact on the atmospheric burden of fluorinated acids. The atmospheric lifetimes of PF-2M3P, PF-3M2B, and 2M3P were estimated to 3-11 days, similar to 13 days, and 1-2 days, respectively. Combining the obtained data, it has been concluded that with 100-year time horizOn global warming potentials (GWP(100)) equivalent to <0.21, similar to 0.29, and <= 1.3 X 10(-7) for PF-2M3P, PF-3M2B, and 2M3P, respectively, these compounds will have a negligible impact on global warming.