pawsplay
Hero
Whether or not it could be a science would depend upon how the magic worked. One can imagine a magical system upon which the scientific method is thoroughly ineffective.
I cannot.
For example - the scientific method depends on the laws governing a phenomenon behaving reproducibly. If I take action X, and get result Y, pretty much every time I take action X, I will get result Y (with margins of error for my not doing *exactly* X each time).
Yet there must be some actions X which are spells, as opposed to all the things that are not.
So, if magic does not act reproducibly, such that effect does not follow clear cause, then science would fail to apply.
It would be impossible to practice. However, science would be necessary to uncover various ways in which to avoid casting horrific spells.
Divine magic, for example - if the operation of magic is dependent on the will of a fickle divine being, then it might well remain impenetrable to science.
Howso? I can think of nothing more interesting to science than the verifiable existence of powerful, reality-bending entities. Simply determining some pattern to their interests, however capricious, would eclipse all other fields of endeavor in importance. What use to build empires, when an inscrutable entity might turn an empire, overnight, into a vast garden of talking turnips? In the face of superenatural superpower, nothing would be more pertinent than science.
The only things impenetrable to science are things that have no causes behind them, which describes... nothing.