Title: The curious case of Mars' formation Abstract: Recent isotopic studies of the martian meteorites suggests that the bulk composition of Mars is significantly different from Earth, with Mars being composed of a higher fraction of ordinary chondrite-like material. Here, we evaluate this compositional difference in more detail by comparing output from two N-body planet formation models: the "Classical" case wherein Jupiter and Saturn are kept in their current orbits, and the "Grand Tack" case in which Jupiter and Saturn migrate through the primordial asteroid belt. Our results show that both models tend to yield Earth and Mars analogues whose accretion zones overlap. Assuming a disc with a composition change (break) at 1.3 AU reproduces the isotopic differences between Earth and Mars in both models. At this break location, the Classical case fares better in forming Mars with its documented composition (29% to 68% enstatite chondrite plus 32% to 67% ordinary chondrite) though the Mars analogues are generally too massive. However, if a restriction of mass on the Mars analogues is included such as requiring that its mass only differ from its current value by at most 50%, the Classical model does not work better. We further calculate the isotopic composition of 17O, 50Ti, 54Cr, 142Nd, 64Ni, and 92Mo in the martian mantle from the Grand Tack simulations. We find that it is possible to match the calculated isotopic composition of all the above elements in Mars' mantle with their measured values, but the resulting uncertainties are too large to place good restrictions on the early dynamical evolution and birth place of Mars. These large uncertainties are caused by both the chaotic evolution of the forming planets and the large uncertainties in the measured isotopic deviations of some elements in martian samples. Hence, increase in both the resolution of N-body simulations and the precision of isotopic measurement from martian meteorites are required to solve the origin of Mars and the other terrestrial planets.