Crushed and homogenized aliquots of samples of basalt and diabase were analysed for 36Ar after various storage times, baking temperatures, and baking times. Il has been found that samples with panicle size greater than 100 mesh, and freshly crushed or stored in vacuum can be baked under vacuum to remove atmospheric argon contamination without loss of radiogenic argon. Using such procedures, initial 36Ar content can be determined uniquely and reproducibly. With removal of atmospheric argon contamination it is thus possible to apply the 40Ar/36Ar vs. 40K/36Ar isochron technique to K–Ar dating in a way exactly analogous to its use in Rb–Sr dating. An attempt is made to clarify some of the causes of misunderstanding that have contributed to improper use of K–Ar isochrons and to set out some of the advantages of the proper use of the method.