We present new models of the magnetic fields of Uranus and Neptune, ba
sed on data provided by the Voyager II magnetic field experiment. We f
ind the simplest models that satisfy the data, and use the observed su
rface heat flow as a constraint on the magnetic field structure. Our m
odels are similar to the previously described Q(3) and O-8 models far
from the planets, but resolve smaller-scale structure close to the pla
nets' surfaces. The field of Neptune is much better constrained than t
hat of Uranus, and the field in the northern hemisphere of Neptune is
much better constrained than that in the southern hemisphere, of impor
tance for studies of Neptunian northern hemisphere radio sources. Usin
g extremal models, we show that the large dipole tilts and nondipole d
ominance of the fields are robust features required by the data. Scali
ng analysis suggests that the toroidal field that would be required fo
r a magnetostrophic balance in the dynamo region would result in ohmic
dissipation greater than the observed surface heat flow. Thus we sugg
est that the dynamos of Uranus and Neptune are energy limited, and tha
t the subsequent lack of magnetostrophic balance may account for the r
adically different field morphologies of Uranus and Neptune compared w
ith the Earth, Jupiter, and Saturn.