The following aspects of buckling tests of shell structures are discussed: (a) the introduction of load into tests specimens without causing undue stress concentrations or deformations, (b) the selection of materials for the construction of test specimens, (c) the use of strain gages in providing information essential to understanding specimen behavior, and (d) the feasibility of experimental studies that attempt to demonstrate the light-weight characteristics of shells whose buckling behavior has not been established. In each instance, the discussion is exemplified by test results from the literature. Several instances are noted in which the buckling behavior of shells is altered considerably by small changes in specimen design and in which the use of strain gages was helpful in understanding specimen behavior.