Review of recent publications indicates that current primary laboratory-calibration methods provide sufficient accuracy for present and foreseeable future needs. An upper frequency limit for vibration measurement was first indicated by Schloss. For usual levels, this sets a lower bound to the amplitudes needed in calibration. The use of resonant systems provides means for reaching large amplitudes as they are needed. Serious errors in calibration result from transverse shake table motion. Methods for monitoring motion exist, allowing elimination of this source of error. Shakers free of transverse motion are needed urgently. Development of means for accurate measurement of the transverse sensitivity of pickups at all frequencies of interest is needed also. Full exploitation of primary calibration techniques requires suitable comparison shakers and transfer standards. Available instruments are described. The need for an adequate transfer standard pickup and the characteristics of such an instrument are discussed. To detect damage in storage and attachment or loss of accuracy caused by structural conditions or in associated circuitry, pre-use calibration checks can be performed in a number of ways. Most available methods separate the pickup from the structure and are limited in frequency and amplitude. A new instrument described by Orlacchio and Schilling allows in-place checking of entire systems over wide ranges of frequency and amplitude. Some shortcomings of the method have been indicated by Pennington and alternatives have been suggested. Published advantages, objections, and inadequacies of each method are discussed. Also, we suggest comparison of the attached pickup with a detachable, laboratory-calibrated, transfer standard, both driven in place by a pulse.