We address the question how accurately stellar ages can be determined
by stellar evolution theory. We select the star with the best observat
ional material available - our Sun. We determine the solar age by fitt
ing solar evolution models to a number of observational quantities inc
luding several obtained from helioseismology, such as photospheric hel
ium abundance or p-mode frequencies. Different cases with respect to t
he number of free parameters and that of the observables to be fitted
are investigated. Age is one of the free parameters determined by the
procedure. We find that the neglect of hydrogen-helium-diffusion leads
to ages deviating by up to 100% from the true, meteoritic solar age.
Our best models including diffusion yield ages by about 10% too high.
The implication for general stellar age determination is that a higher
accuracy than that can not be expected, even with the most up-to-date
models. Our results also confirm that diffusion as treated presently
in solar models is slightly too effective.