Inorganic arsenic (iAs) has been classified as a type 1 carcinogen and has also been linked to several noncancerous health effects. Prior to 1995, the As(V) methylation pathway was generally considered to be a detoxification pathway, but cellular and animal studies involving MMA(III) (mono metyl arsonous acid) and DMA(III) (dimethyl arsinous acid) have indicated that their toxicities meet or exceed that of iAs, suggesting an activation process. In addition, thiolated arsenic metabolites were observed in urine after oral exposure of inorganic arsenic in some studies, for which the toxicological profile was not yet fully characterized in human cells. Studies have revealed that microorganisms from the gut environment are important contributors to arsenic speciation changes. This presystemic metabolism necessitates the development of protocols that enable the detection of not only inorganic arsenic species, but also pentavalent and trivalent methylated, thiolated arsenicals in a gastrointestinal environment. We aim to study the biotransformation of arsenic (As) using a Simulator of the Human Intestinal Microbial Ecosystem (SHIME). To be able to analyze the arsenicals resulting from biotransformation reactions occurring in this system, a method using liquid chromatography hyphenated to an inductively coupled plasma mass spectrometer (HPLC-ICP-MS) was developed. A Hamilton PRP-X100 anion exchange column was used. The method allowed separation, identification and quantification of As(III) (arsenite), As(V) (arsenate), DMA(V) (dimethylarsinicacid), MMA(V) (monomethylarsonicacid) and MMMTA (monomethylmonothioarsenate). Attempts to optimize the same method for also separating MMA(III) and DMA(III) did not succeed. These compounds could be successfully separated using a method based on the use of a Zorbax C₁₈ column. The properties of the column, buffer strength, pH and polar nature of mobile phase were monitored and changed to optimize the developed methods. Linearity, sensitivity, precision, accuracy and resolution of both methods were checked. The combination of the two methods allowed successful quantification of arsenic species in suspensions sampled in vitro from the SHIME reactor or in vivo from the human colon and feces.