Rammed earth possesses environmental advantages over most other competing construction materials. However, if it is to be more routinely used in the construction of modern, sustainable buildings, its material properties and production processes must be properly quantified. This paper proposes practical recommendations for soil selection, stabilizer treatment, and on-site compaction for rammed earth, based on a recent set of 219 stabilization experiments. The purpose of the recommendations is to maximize the probability of constructing rammed earth walls that meet or exceed a compressive strength criterion of 2 MPa. The recommendations cover: (1) Quantifying the natural soil properties of linear shrinkage and texture in a staged sequence in order to identify suitable soils to stabilize (and to reject unsuitable soils); (2) Quantifying the amounts of cement and/or lime to be added to the selected soil according to the values of soil properties measured; and (3) Quantifying the forces involved in on-site compaction of stabilized soil (for both manual and pneumatic ramming), and relating these to laboratory-based test standards. Although the recommendations need to be tested and verified/refined using new data, their initial application to rammed earth construction situations in Australia indicates that they have predictive utility. Further research will also indicate the degree of applicability of the recommendations to the production of compressed earth bricks.