Previous attempts to determine the distribution of ice in frozen tissues at high sub-zero temperatures generally called for the further cooling of the tissues in question to facilitate freeze-drying, freeze-substitution, and freeze-fracture replication. Direct cryomicroscopic determinations, free from uncertainties stemming from changes in sample temperature could, it seemed, only be made in certain special cases. We have presented an isothermal “freeze-fixation” procedure designed to permit, instead, the postthaw retention of the freezing pattern and the conventional processing, afterward, of the thawed specimen. The method demands the exposure of the frozen tissues to fixative solutions incapable of dissolving ice. Frozen specimens are immersed in aqueous fixative solutions prepared in each instance (1) to freeze at a temperature equal to that at which fixation is to be conducted, (2) to contain quantities of finely divided ice sufficient to guarantee the maintenance of a constant water activity. Frozen frog and rat hearts and skeletal muscle tissues were exposed to formaldehyde, formaldehyde/ glutaraldehyde, and glutaraldehyde solutions at −2, −5, and −10 °C, the temperatures being maintained in each case to ± 0.1 °C, or better. Tissues withdrawn at intervals were thawed, postfixed, dehydrated, embedded, and sectioned. The sections demonstrated the retention, after thawing, of structural features characteristic of the frozen state. The small hearts we exposed to formaldehyde were fixed throughout in 3 hr at −2 ° and in 20 hr at −5 °C. The action of osmium tetroxide was investigated. The method appears to be well-suited to numerous experimental applications.