Three quite different things share this word and running them together is the commonest error in the subject.
The first is the European witch trials of roughly 1450 to 1750. Tens of thousands of people were executed, most of them women, on charges assembled by prosecutors from a mixture of learned demonology and local suspicion. The historical consensus is that the crime as charged did not exist and that the confessions were produced by the process. That is a finding about the trials, not about whether anyone practised anything.
The second is the folk practice that did exist alongside them: cunning folk, charmers and healers who cured, found lost property and lifted curses, and who were often the people villagers went to for protection against witchcraft. Their trade is well attested and it was not the thing being prosecuted, though practitioners were sometimes caught up in it.
The third is modern Wicca. Margaret Murray argued in the 1920s that the trials had persecuted a surviving pre-Christian religion, a thesis historians have comprehensively rejected. Gerald Gardner nevertheless built on it, and after the repeal of the Witchcraft Act in Britain in 1951 published an account of a religion he presented as that survival. Its elements are traceable to ceremonial magic, folklore and Gardner's own composition.
Wicca is a genuine modern religion with real adherents. Its claim to descend from the people burned in the trials does not hold.