This paper studies a behavioral model of persuasion in which the receiver probabilistically misinterprets the sender-designed information policy. Such perturbation unravels the concavification technique, but we preserve belief approach by attaching misinterpretation probabilities to induced beliefs. We show that misinterpretation reduces both the sender’s optimal persuasion value and the total communication surplus via bounded implementability. The receiver remains unaffected so long as Bayesian updating with respect to the effective information environment. In the binary case, we solve for the sender's optimal value from misinterpreted persuasion. We also looked at when the receiver is unaware of misinterpretation and analyzed interaction between misinterpretation and this naïve misspecification. We find that if there is misinterpretation favoring the sender, receiver's naïveté impairs optimal decision-making, benefiting the sender by reducing the receiver's equilibrium demand for information. Finally, we extend the binary case to incorporate confirmation bias, where the direction of misinterpretation is endogenously determined.