The changes in the aortal pulse-wave velocity (PWV) occurring in connection with and dependent on spontaneous fluctuations of blood pressure has been recorded at intervals over the years in 183 normotensive men, who were aged from 17-74 at the beginning of the long term study. An essential condition was that the diastolic blood pressure should not exceed 95 mm Hg and the systolic 145 mm Hg. In order to eliminate the effect of age on the PWV between 2 measurements, and so to obtain a "pure" PWV-mean pressure relationship (c-p relationships), 7 cm/sec per year was subtracted from the c-value of the second measurement before the age of 55 and 9 cm/sec per year after that age. The differences quotient delta cp as a standard for the change of c with the mean pressure p was obtained from the difference between two c values (c1-c2) taken at different times, converted to a pressure change of 10 mm Hg, and divided by the difference of the mean pressure levels belonging to them (p1-p2). In applying the appropriate age correction to c2, the time factor had no statistically recognisable effect on delta cp. In 78% of the cases in our long term study, c rose and fell with p, in 22%, the changes of c were at variance with the changes of p. Taking into consideration all the test subjects, delta cp averaged 0.40 m/sec. Before the age of 55, delta cp is smaller (0.30 m/sec) than above that age (0.55 m/sec). The age difference of delta cp is significant ( = 0.05). When the concordant c-p relationships alone were calculated, delta cp was 0.70 m/sec and scarcely differed from the delta cp values of hypertensives published earlier (0.60 m/sec: also concordant c-p relationships only). The generally lower delta cp values from group cross sections (in contrast to the longitudinal investigations) are explained by an unrecognisable admixture of discordant c-p relationships. In a range of pressure from 90-170 mm Hg, delta cp was shown to be independent of the level of the initial pressure. Also the magnitude of the (spontaneous) mean pressure variation (5-70 mm Hg, normotensives - hypertensives) seems to have no effect on the statistical mean value of delta cp. delta cp is, however, dependent on the direction of the pressure change in normotensives (just as with hypertensives), even when age is taken into account. If the pressure is reduced, c is higher and delta cp (p = 0.05) is greater than when the blood pressure is increased. The c-p relationship traverses a kind of loop (counterclockwise). In the discussion, an attempt is made to point out the effect of the vascular musculature on delta cp, which threads conspicuously through the comparison of the physiological delta cp values with the delta cp values in arteriosclerosis (hypertension; diabetes) and in endurance training. From this it can be deduced that normal values for c and delta cp in arteriosclerosis indicate that the musculature is still capable of maintaining a normal elastic function even with considerable regressive changes in the vessel wall.