Electron probe energy dispersive x-ray microanalysis (XRMA) offers a powerful tool for the determination of intracellular elemental content of biological tissue. However, preparation of the tissue specimen , particularly excitable central nervous system (CNS) tissue , for XRMA is rather difficult, as dissection of a sample from the intact organism frequently results in artefacts in elemental distribution. To circumvent the problems inherent in the in vivo preparation, we turned to an in vitro preparation of astrocytes grown in tissue culture. However, preparations of in vitro samples offer a new and unique set of problems. Generally, cultured cells, growing in monolayer, must be harvested by either mechanical or enzymatic procedures, resulting in variable degrees of damage to the cells and compromised intracel1ular elemental distribution. The ultimate objective is to process and analyze unperturbed cells. With the objective of sparing others from some of the same efforts, we are reporting the considerable difficulties we have encountered in attempting to prepare astrocytes for XRMA.Tissue cultures of astrocytes from newborn C57 mice or Sprague Dawley rats were prepared and cultured by standard techniques, usually in T25 flasks, except as noted differently on Cytodex beads or on gelatin. After different preparative procedures, all samples were frozen on brass pins in liquid propane, stored in liquid nitrogen, cryosectioned (0.1 μm), freeze dried, and microanalyzed as previously reported.