Xceleron Inc., 20340 Seneca Meadows Parkway, Germantown, MD 20876, USA Tel.: +44 142 353 8423 E-mail: mark.seymour@ xceleron.com Accelerator MS (AMS) was first applied in the biomedical field in the late 1990s and was initially used principally in human metabolism studies where the sensitivity of conventional radioactivity detectors was insufficient, or in human Phase 0 microdosing. The key driver for using AMS for microdosing was again sensitivity, which enabled measurement of circulating concentrations arising from doses of 100 μg or less. However, despite early predictions that microdosing would engender a paradigm shift in drug development, continuing concern regarding the predictivity of the PK data generated coupled with the continuous and continuing improvements in the sensitivity of LC–MS/MS assays has limited adoption of the approach. Furthermore, while AMS continues to be used for metabolism studies, the advantages of reduced radiation exposure to subjects and the potential to administer multiple radiolabeled doses has not driven widespread adoption. Instead, over the last 5 years, AMS has been increasingly used in enhanced Phase I PK studies, where the ability of the detector to discriminate between C-labeled and nonlabeled doses facilitates an innovative clinical study design whereby a low dose (typically 10 μg or less) of C-compound is administered intravenously contemporaneously with a pharmacologically relevant extravascular dose of the nonradiolabeled compound. This so-called IV/PK, or absolute bioavailability, study design has a number of important advantages over the conventional intravenous (i.v.)/extravascular crossover design. Scientifically, the IV/PK design is optimal: since both doses are administered at the same time there can be no variation due to temporal effects and the compound administered by both dose routes is handled by the body in exactly the same way, so the potential effect of nondose proportional kinetics does not have to be considered. Notwithstanding the latter point, the actual dose administered by the i.v. route is so low that no i.v. safety toxicology studies are required prior to human dosing. Additionally, the task of developing a suitable i.v. formulation is facilitated, particularly for poorly soluble compounds, with a simple aqueous solution typically used. While the utility of the IV/PK design is widely accepted, there is an ongoing debate as to the relative merits of the C-AMS approach compared with labeling the i.v. dose with C and analyzing using LC–MS/MS. It is undoubtedly attractive to be able to work with a stable isotope (rather than radioactive C) and to use essentially the same analytical methodology to quantify compound administered by both dose routes (discrimination being possible through the appropriate selection of ions/transitions to be monitored). On the other hand, C-labeled analogues are often used as internal standards (ISs) in LC–MS/MS assays, so a second analogue must be synthesized and the assay method must be able to distinguish between all three analogues. The sensitivity of the LC–MS/MS assay for the C-analyte is also a potential issue. It is essential that systemic concentrations arising from the i.v. dose are sufficiently low relative to those arising from the extravascular dose to avoid any possibility of perturbing the kinetics of the extravascular dose (at least a 100-fold excess should be the target). This may challenge the LLOQ of the LC–MS/MS assay, particularly for compounds with low bioavailability and/or