SUMMARY1. Observations were made upon five subjects who flew through 4j-6 time zones, four of them returning later to their starting point, and upon twenty-three subjects experiencing simulated 6 or 8 hr time zones shifts in either direction in an isolation unit.2. Measurements were made of plasma concentration of li-hydroxycorticosteroids, of body temperature, and of urinary excretion of sodium, potassium and chloride. Their rhythm was defined, where possible, by fitting a sine curve of period 24 hr to each separate 24-hr stretch of data and computing the acrophase, or maximum predicted by the sine curve.3. The adaptation of the plasma steroid rhythm was assessed by the presence of a sharp fall in concentration after the sample collected around 08.00 hr. The time course of adaptation varied widely between individuals; it was usually largely complete by the fourth day after westward, and rather later after eastward, flights. After time shift the pattern often corresponded neither to an adapted nor to an unadapted one, and in a subject followed for many months after a real flight a normal amplitude only appeared 2-3 months after flight.4. Temperature rhythm adapted by a movement of the acrophase, without change in amplitude, although on some days no rhythm could be observed. This movement was always substantial even on the first day, and was usually nearly complete by the fifth.5. High nocturnal excretion of electrolyte was often seen in the early days after time shift, more notably after simulated westward flights. Adaptation of urinary electrolyte rhythms usually proceeded as with temperature, but the movement of the acrophase was slower, more variable ANN L. ELLIOTT AND OTHERS between individuals, more erratic, and sometimes reversed after partial adaptation. On a few days there were two maxima corresponding to those expected on real and on experimental time.6. Sodium excretion was much less regular than that of potassium, but adapted more rapidly to time shift, so that the two often became completely dissociated. Chloride behaved much as sodium.7. The time course of adaptation of the plasma steroid and urinary potassium rhythms were sufficiently similar to suggest a causal connexion. The time course of adaptation of the temperature rhythm did not coincide with that of any other component considered here.