The development of bacteria-specific infection radiotracers is of considerable interest to improve diagnostic accuracy and enabling therapy monitoring. The aim of this study was to determine if the previously reported radiolabelled 1,4,7,10-tetraazacyclododecane-N,N′,N″,N‴-tetraacetic acid (DOTA) conjugated peptide [⁶⁸Ga]Ga-DOTA-K-A9 could detect a staphylococcal infection in vivo and distinguish it from aseptic inflammation. An optimized [⁶⁸Ga]Ga-DOTA-K-A9 synthesis omitting the use of acetone was developed, yielding 93 ± 0.9% radiochemical purity. The in vivo infection binding specificity of [⁶⁸Ga]Ga-DOTA-K-A9 was evaluated by micro positron emission tomography/magnetic resonance imaging of 15 mice with either subcutaneous Staphylococcus aureus infection or turpentine-induced inflammation and compared with 2-deoxy-2-[¹⁸F]fluoro-D-glucose ([¹⁸F]FDG). The scans showed that [⁶⁸Ga]Ga-DOTA-K-A9 accumulated in all the infected mice at injected doses ≥3.6 MBq. However, the tracer was not found to be selective towards infection, since the [⁶⁸Ga]Ga-DOTA-K-A9 also accumulated in mice with inflammation. In a concurrent in vitro binding evaluation performed with a 5-carboxytetramethylrhodamine (TAMRA) fluorescence analogue of the peptide, TAMRA-K-A9, the microscopy results suggested that TAMRA-K-A9 bound to an intracellular epitope and therefore preferentially targeted dead bacteria. Thus, the [⁶⁸Ga]Ga-DOTA-K-A9 uptake observed in vivo is presumably a combination of local hyperemia, vascular leakiness and/or binding to an epitope present in dead bacteria.