The paper continues the search for general rules governing the electron density distribution in organic molecules and molecular systems on the basis of application of the non-canonical method of molecular orbitals. It is devoted to investigation of an electron-donating effect of an external orbital upon a single (Z–C) or a double (ZC) heteroatom-containing bond, which is among the widespread effects of organic chemistry (cf. a nucleophilic attack upon substituted alkanes or carbonyl compounds). Use of expressions for occupation numbers of basis orbitals of a Z–C(ZC) bond derived previously [V. Gineityte, J. Mol. Struct. (Theochem) 434 (1998) 43] allows the additional dipole moment of the bond due to the electron-donating effect to be considered as a result of competition between two meaningful components of opposite signs, viz. between the secondary polarization of the given bond owing to the very presence of the electron-donating orbital and the so-called depolarization originating from the dipole-like distribution of the acquired population. The consequent direction of the additional dipole is shown to depend decisively on the relative electronegativity of the heteroatom Z and thereby on the initial polarity of the bond. Predominance of the secondary polarization and thereby coincidence between the directions of the primary dipole of the given bond and of the additional one is obtained for bonds of relatively low initial polarity but not for those of high polarity. This somewhat unexpected result is traced back to alterations in constitution of the antibonding orbital of the Z–C(ZC) bond when the electronegativity of the heteroatom Z grows that yield a drastic increase of the absolute value of the depolarization dipole. On the basis of the results obtained, the rule of the so-called ‘curly arrow chemistry’ about the universal increase of the dipole moment of a Z–C(ZC) bond under influence of an electron-donating effect is concluded to be of a limited scope of validity. Moreover, a new accounting is suggested for the well-known experimental fact that highly electronegative heteroatoms usually are bad nucleofuges in S N 2 processes.