The slow expansion of the bright ring around SN 1987A and the large si
ze estimated for the radio source at an age of similar to 1500 days su
ggest a low mass-loss rate, M, from the blue supergiant progenitor sta
r; we adopt M = 7.5 x 10(-8) M(.) yr(-1) and a wind velocity upsilon(w
) = 450 km s(-1) as standard values. This wind density would not give
substantial radio free-free absorption during the first few days when
low-frequency absorption was observed, but the turnover can be attribu
ted to synchrotron self-absorption. The X-ray and radio emission obser
ved since 1990 require that the supernova shock be interacting with de
nser gas than that in the free wind. We propose that it is interacting
with an H II region created by the B3 Ia progenitor star in the swept
-up red supergiant wind. The H II region extends from a radius r(II) a
pproximate to 3 x 10(17) cm to ring radius R = 6 x 10(17) cm in the eq
uatorial plane and has a hydrogen density similar to 10(2) cm(-3). The
shock front will not reach the dense ring until the year 2005 +/- 3.
The X-ray luminosity should slowly increase until that time. The equat
orial ionization front is trapped in the swept-up gas; the dense ring
and its associated bipolar nebula are the initially neutral part of th
e shell. In the polar direction, the ionization front broke out of the
swept-up gas, and there is no dense neutral shell, consistent with th
e observed lack of a thin shell in that direction.