Over the years, there have been many attempts to delineate the boundary between magic and religion, but this cannot be done. It was once a common practice among scholars writing about "magic" to distinguish it from "religion" and to treat them as somewhat exclusive of one another. However, as has been convincingly argued by more recent scholars, the various forms of magic must be viewed as a subset of religious practices – practices that were outside standard cult rituals, and, though they often involved exotic divinities or even spirits whose names only occur in magical contexts, were nevertheless religious in essence. Magic was a means for achieving ends that its practitioners felt could not be reached with certainty through ordinary prayers and rituals, but like traditional cults it too relied on achieving a close working relationship with supernatural powers. There are numerous other similarities between ancient "magic" and "religion": both often involved the performance of rituals or recitations of elaborate hymns or prayers, and it was common for one activity to accompany the other; they shared many of the same goals, especially maintaining or regaining one’s health, as well as seeking divine advisories on whether one should proceed with a certain venture, or attempting to obtain a favorable outcome; both relied on established bodies of knowledge and, quite often, experts who maintained such knowledge and used it to assist others; and, despite occasional innovations, both were extremely traditional. Perhaps the most significant distinguishing characteristic was that magic was performed privately, individually, and often in secret. Cult practices, on the other hand, often took place publicly in temples or during festivals, and even ceremonies of the secretive mystery cults were conducted by groups of worshippers rather than lone individuals. However, there is also abundant evidence for individuals who sought direct and personal relationships with their chosen gods by worshipping them in private, so the setting of magical practices turns out to be yet another insufficient distinction. In truth, any attempt to draw a clear line between magical and religious practices will inevitably fail, since there will always be an exception or parallel that rules out any firm conclusions.

If we are to be able to distinguish "magic" from mainstream "religion," then, our best option is to consider the attitude and goals of the individual practitioner. In other words, we should pay attention not to what was done, but why these things were done. Generally speaking, traditional cult practices involved attempts to gain the favor of a god, and this was usually done by making offerings as payment for their recent assistance or promising future offerings in return for the god’s support. The worshipper approached the god as the inferior partner in the relationship, although he or she did have some degree of leverage in that the god presumably desired the offering as well as the attention. Those who performed magic sought a very different relationship, one in which the magic-users were dominant. Despite the dizzying variety of spells and practices, the fundamental principle of most magic was the desire to gain control over a divinity and, having enslaved it, compel it to serve the practitioner. Thus magical papyri from Egypt give instructions on how to obtain a spirit-servant, while curse tablets from North Africa and the rest of the Mediterranean basin include threats and commands aimed at divinities, rather than prayers and entreaties. This attempt to distinguish between "magic" and "religion" based on the practitioner’s mindset, however, is admittedly flawed, since it applies more to aggressive magic than the much more prevalent forms of benevolent, or "white," magic. A person who was creating a lucky charm or an amulet to prevent fever might have addressed it to a specific divinity, but the surviving texts of such items do not use any of the assertive language found in spell-books and curse tablets. When the texts of amulets and other forms of beneficial magic named deities, it was in the context of enlisting divine support instead of seeking to compel it. Such an attitude seems quite similar to that inherent in ordinary religious practices, but even in antiquity amulets were judged to be extra-ordinary. It seems most reasonable, then, to conclude that ancient "magic" should not be viewed as monolithic and, furthermore, we should not ascribe a single mindset to all of its practitioners. In the end, the relation between "religion" and "magic" can best be understood in terms of degrees: the attitudes inherent in the most aggressive and hostile forms of magic had the least in common with those of mainstream religion, while forms of magic meant to help the user without bringing harm to others were so close in nature to ordinary religious attitudes as to be virtually indistinguishable.