In order to test the accuracy of the Heavy Top-mass Expansion (HTE) employe
d in recent two-loop calculations of M-w and sin(2) theta(eff)(lept), we co
nsider their contributions to subtracted quantities of the form (M-w)(sub)
= M-w(M-H) - M-w(M-H(o)) where M-H(o) is a reference point. The results are
compared with those obtained by a precise numerical evaluation of all the
two-loop contributions involving both the Higgs boson and a fermion loop. F
or the choice M-H(o) = 65 GeV, and over the large range 65 GeV less than or
equal to M-H less than or equal to 1 TeV, we find very small differences b
etween the precise and HTE calculations, amounting to \delta M-w\ less than
or equal to 0.8 MeV and \delta s(eff)(2)\ less than or equal to 1.2 x 10(-
5). Although corrections involving light fermions are necessary for the con
sistency and test of existent calculations, we also discuss the separate co
ntributions from the top-bottom isodoublet. In this case, the differences a
re larger, although still small, namely \delta M-w\ less than or equal to 1
.9 MeV and \delta s(eff)(2)\ less than or equal to 4.5 x 10(-5).