The early development of Mars is of enormous interest, not just in its own right, but also because it provides unique insights into the earliest history of the Earth, a planet whose origins have been all but obliterated. Mars is not as depleted in moderately volatile elements as are other terrestrial planets. Judging by the data for Martian meteorites it has Rb/Sr ~ 0.07 and K/U ~ 19,000, both of which are roughly twice as high as the values for the Earth. The mantle of Mars is also twice as rich in Fe as the mantle of the Earth, the Martian core being small (~20% by mass). This is thought to be because conditions were more oxidizing during core formation. For the same reason a number of elements that are moderately siderophile on Earth such as P, Mn, Cr and W, are more lithophile on Mars. The very different apparent behavior of high field strength (HFS) elements in Martian magmas compared to terrestrial basalts and eucrites may be related to this higher phosphorus content. The highly siderophile element abundance patterns have been interpreted as reflecting strong partitioning during core formation in a magma ocean environment with little if any late veneer. Oxygen isotope data provide evidence for the relative proportions of chondritic components that were accreted to form Mars. However, the amount of volatile element depletion predicted from these models does not match that observed - Mars would be expected to be more depleted in volatiles than the Earth. The easiest way to reconcile these data is for the Earth to have lost a fraction of its moderately volatile elements during late accretionary events, such as giant impacts. This might also explain the non-chondritic Si/Mg ratio of the silicate portion of the Earth. The lower density of Mars is consistent with this interpretation, as are isotopic data. ^87Rb-^87Sr, ^129I-^129Xe, ^146Sm-^142Nd, ^182Hf-^182W, ^187Re-^187Os, ^235U-^207Pb and ^238U-^206Pb isotopic data for Martian meteorites all provide evidence that Mars accreted rapidly and at an early stage differentiated into atmosphere, mantle and core. Variations in heavy xenon isotopes have proved complicated to interpret in terms of ^244Pu decay and timing because of fractionation thought to be caused by hydrodynamic escape. There are, as yet, no resolvable isotopic heterogeneities identified in Martian meteorites resulting from ^92Nb decay to ^92Zr, consistent with the paucity of perovskite in the martian interior and its probable absence from any Martian magma ocean. Similarly the longer-lived ^176Lu-^176Hf system also preserves little record of early differentiation. In contrast W isotope data, Ba/W and time-integrated Re/Os ratios of Martian meteorites provide powerful evidence that the mantle retains remarkably early heterogeneities that are vestiges of core metal segregation processes that occurred within the first 20 Myr of the Solar System. Despite this evidence for rapid accretion and differentiation, there is no evidence that Mars grew more quickly than the Earth at an equivalent size. Mars appears to have just stopped growing earlier because it did not undergo late stage (>20 Myr), impacts on the scale of the Moon-forming Giant Impact that affected the Earth.