Following the passage of the Waxman-Hatch Act (1984), FDA approval for a generic drug requires the establishment of bio-equivalence between the generic drug and an FDA approved branded drug. However, a large body of evidence in the medical community suggests that bio-equivalence does not guarantee therapeutic equivalence; in some instances the lack of therapeutic equivalence can lead to fatal consequences for patients switching to generic products. In this paper, we construct a simple model to analyze the implications of therapeutic non-equivalence between branded and generic drugs. We show, theoretically and empirically, that this distinction can provide a plausible explanation of the generic competition paradox.