Antibiotics are released into the environment in a variety of ways: via wastewater effluent as a result of incomplete metabolism in the body after use in human therapy, as runoff after use in agriculture, through improper disposal by private households or hospitals or through insufficient removal by water treatment plants. Unlike in most European countries, in Arctic regions effluents are not suitably treated prior to their release into the aquatic environment. Also, many of the scattered human settlements in remote regions of the Arctic do not possess sewage treatment facilities and pharmaceutical residues therefore enter the aqueous environment untreated. Only limited data are available on the biodegradation of antibiotics under Arctic conditions. However, such information is needed to estimate the potential harm of antibiotics for the environment. Pen-G is used in this study since it is a widely prescribed antibiotic compound whose environmental properties have not yet been investigated in detail. Thus, for a very first assessment, the OECD approved biodegradation Zahn-Wellens test (ZWT, OECD 302 B) was used to study biodegradation and non-biotic elimination of the antibiotic Benzyl-penicillin (Pen-G) at different temperatures (5 °C, 12.5 °C and 20 °C). The testing period was extended from the OECD standard of 28-42. d. In addition to dissolved organic carbon (DOC), Pen-G levels and major transformation products were recorded continuously by LC-ion-trap-MS/MS. DOC monitoring revealed considerable temperature dependence for the degradation process of Pen-G. DOC loss was slowest at 5 °C and considerably faster at 12.5 °C and 20 °C. In the initial step of degradation it was found that Pen-G was hydrolyzed. This hydrolyzed Pen-G was subsequently further degraded by decarboxylation, the result of which was 2-(5,5-dimethyl-1,3-thiazolidin-2-yl)-2-(2-phenylacetamido)acetic acid. Furthermore, direct elimination of 2-phenyl-acetaldehyde from the hydrolyzed and decarboxylated Pen-G also led to the formation of 2-[amino(carboxy)methyl]-5,5-dimethyl-1,3-thiazolidone-4-carboxylic acid. Since biodegradation slows down considerably at a low temperature, the resulting transformation products had considerably longer residence times at 5 °C compared to higher temperature conditions within the 42-d experiment. The results presented here clearly demonstrate that a risk assessment for pharmaceuticals present in low ambient temperature environments (i.e. the Arctic) cannot be based on test results obtained under standard laboratory conditions (i.e. 20 °C ambient temperatures). © 2010.