A model is presented for the global thermoremanent magnetization of spherical lithospheres which cool in the presence of central dipole fields. Reversals and intensity variations of the field are incorporated in this model, which is applicable to bodies whose interiors have remained above the Curie point throughout their evolution. The model demonstrates that even considering Runcorn's magnetostatistics theorems for spherical shells, a nonzero magnetic permeability and a finite cooling rate in the lithosphere permit the acquisition of a sizable global remanent dipole moment, which would be detectable by external measurements after the magnetizing field has disappeared. Preliminary application of this model to Mercury, Venus, and Mars suggests that only the combination of a nonreversing ancient source field with a surface value near 1 Oe plus a sizable concentration (~1% by volume) of ferromagnetic material in their crusts could produce remanent planetary dipole fields as large as those measured by spacecraft. On the other hand, if ancient reversing dynamos existed in these planets, it is unlikely that large planetary-scale fields like those observed at Mercury could be due to remanence in their crusts, irrespective of their composition.