We have calculated a large number of linear radial pulsation models to
fit the periods of AC And. In the parameter regime representative of
RR Lyrae stars no models can be found that match the observed periods
for any heavy element abundance Z. There are, however, two groups of h
igher mass models which yield complete agreement between the observed
periods and those of the first three radial modes. The first comprises
models with 0.001 < Z < 0.002 and masses and luminosities of 1.1 < M/
M. < 1.4, 100 < L/L. < 130 (depending on Z). The second group has high
er masses, luminosities and Z, viz. 0.01 < Z < 0.02, M/M. almost-equal
-to 2.7, 80 < L/L. < 200. These ranges are sensitive to the assumed un
certainties of the model periods. We give ranges of model parameters a
s functions of Z and discuss their uncertainties. Present observationa
l constraints and the limited sensitivity of the method of fitting per
iods do not allow us to sufficiently narrow down the derived parameter
ranges. Furthermore, although recent evolution calculations are in fa
ir agreement with both low- and high-mass models, the latter ones are
preferred because the theoretical period changes are in better agreeme
nt with the observations. Therefore, we suggest that AC And is an evol
ved high-mass star, with similarities to the delta Scuti type stars, o
r perhaps a rare oddball. As a side product of our survey, we give acc
urate fitting formulae for the periods and period ratios.