The problem of measuring the solar age by means of helioseismology has been recently revisited by Guenther & Demarque ( 1997 ) and by Weiss & Schlattl ( 1998 ) .
Different best values for t _ { seis } and different assessment of the uncertainty resulted from these two works .
We show that depending on the way seismic data are used , one may obtain t _ { seis } \approx 4.6 Gy , close to the age of the oldest meteorites , t _ { met } = 4.57 Gy , like in the first paper , or above 5 Gy like in the second paper .
The discrepancy in the seismic estimates of the solar age may be eliminated by assuming higher than the standard metal abundance and/or an upward revision of the opacities in the solar radiative interior .

We argue that the most accurate and robust seismic measure of the solar age are the small frequency separations , D _ { \ell,n } = \nu _ { l,n } - \nu _ { \ell + 1 ,n - 1 } , for spherical harmonic degrees \ell = 0 , 2 and radial orders n \gg \ell .
The seismic age inferred by minimization of the sum of squared differences between the model and the solar small separations is t _ { seis } = 4.66 \pm 0.11 , a number consistent with meteoritic data .
Our analysis supports earlier suggestions of using small frequency separations as stellar age indicators .