Magnetotactic bacteria (MTB) produce intracellular magnetite (Fe3O4) nanoparticles in a genetically controlled manner. They may represent some of the oldest biomineralizing organisms available in the geological record, but identification of their fossils remains highly debated. While organic molecules are degraded during diagenesis and metamorphic processes, MTB magnetite nanocrystals can be efficiently preserved in the rock record and are referred to as magnetofossils. Experimental work on the freshwater bacterium Magnetospirillum magneticum strain AMB-1 has demonstrated specific minor and trace element patterns distinct from those of abiotic magnetite, and were proposed as a tool for magnetofossil identification. These promising geochemical signatures need to be validated in diverse MTB strains to be used for paleontological reconstruction. Here, we cultivated a marine MTB (Magnetovibrio blakemorei strain MV-1) under various chemical conditions to test possible generalization of this new proxy. MV-1 was grown under various Fe concentrations (50, 100 and 150 μM) and redox states using either Fe(II)-ascorbate or Fe(III)-citrate as Fe sources. The chemical compositions of the growth media and extracted magnetite crystals were determined by ICP-MS analyses to quantify the partitioning of trace and minor elements between magnetite and solution. Results show that partition coefficients do not depend at first order on the Fe concentration and redox state, a crucial conclusion for potential application to natural systems. A comparison of the two strains shows that MV-1 magnetite generally contains higher concentrations of impurities than AMB-1 magnetite. However, a number of elements possess similar partition coefficients and may represent useful chemical proxies for testing the biological origin of magnetite. These consistent elements can be separated into three groups. The first group is composed of elements (Co, Mn, Pb, Sr) highly depleted in MTB magnetite relative to abiotic magnetite. The second group contains elements with similar partitioning in MTB and abiotic magnetite, including Ca and Li. This group may serve as a reference for constraining a paleo-fluid composition. The last group contains elements (Mo, Sn, Se) enriched in MTB magnetite relative to abiotic magnetite. Such enrichments might be related to biological function of those elements. Chemical patterns determined from laboratory experiments therefore represent promising chemical proxies to identify MTB magnetite in the rock record but now need to be tested in modern natural environments, where MTB and surrounding solution can be jointly collected.