Martian meteorites have provided a wealth of information about martian petrologic evolution, some clues to the nature of the martian hydrosphere and atmosphere and important clues in the search for life on Mars. There are, however some major limitations in using meteorites to model Mars' history.

Martian meteorites are from uncertain, random locations, not sites selected to answer major questions about Mars' history. Ideally we need samples from at least two known locations, one on old cratered highlands and one on young volcanic plains. Analyses of these samples on Earth would provide ground truth for remote sensing and calibrate the martian crater-age trend.

The second limitation is that martian meteorites are not representative of the martian surface. They are subsurface igneous rocks, not surface soils or sediments, and are thus not ideal as ground truth. They also represent extremes of ages of martian geologic provinces - ancient (4.5 Ga) highlands and very young (1.3 Ga or less) basalts and cumulates. We would like samples that are spread through martian history, especially samples bracketing the time of purported martian climate change. Soils and sediments, rather than igneous rocks, are needed to better understand martian weathering and atmospheric evolution and to search for prebiotic chemistry and possible fossil life.

To better understand Mars' geologic history we need returned samples from two or more carefully selected locations. Samples of soil and rocks returned by intelligent robotic rovers could significantly improve our knowledge of Mars. But to really solve the problem we may have to wait for humans to walk on mars and use our acute visual and mental systems to find the best samples.