As part of a working definition of a new code of practice, this paper develops a methodology to determine the suitability of plasters manufactured on-site to protect the earthen walls of vernacular architecture buildings. Given the diversity of raw earth construction types, ranging from massive earth to stone masonry with earth mortars, and the variability of the materials used, two on-site tests (a shrinkage test followed by a shear test) were proposed. Those tests, as well as additional tests, were performed with lime/sand, earth/sand and earth with plaster admixtures. Lime/sand plasters do not typically pose shrinkage issues, which is why more earth based specimens were tested than lime based specimens. An analysis of the on-site testing complemented with laboratory tests reveals an antagonistic control of the bond between earth plaster and earthen wall based on clay content: an increase in the plaster clay content leads to increased bending strength, which strengthens the plaster and increased shrinkage, which weakens the plaster-wall interface. The heterogeneity of the wall leads to a wide range of results; therefore, this study was conducted to validate the formulation of plasters by means of shrinkage and shear tests at five different points on the wall. The shrinkage test allows finding the earth plaster formulations. Among all the validated formulations, the mason chooses the best one thanks to two criteria: the best workability (which is variable according to masons) with the highest clay content as possible. Then, when the formulation is chosen, the shear test must be done to verify if the bond between the plaster and the wall is high enough. This series of tests allows masons to validate formulations that will ensure good mechanical resistance of the plasters that protect buildings of vernacular earthen architecture.