Augmented patient–practitioner interactions that enhance therapeutic alliance can increase the placebo effect to sham treatment. Little is known, however, about the effect of these interactions on maladaptive health outcomes (i.e., the nocebo effect). Healthy participants (N = 84) were randomised to a 3-day course of Oxytocin nasal drops (actually, sham treatment) in conjunction with a high-warmth interaction (Oxy-HW: N = 28), a low-warmth interaction (Oxy-LW: N = 28) or to a no treatment control group (NT: N = 28). All participants were informed that the Oxytocin treatment could increase psychological well-being but was associated with several potential side effects. Treatment-related side effects, unwarned symptoms, and psychological well-being were measured at baseline and all post-treatment days. Side effect reporting was increased in the Oxy-LW condition compared to the other groups across all days. Conversely, increased psychological well-being was observed in the Oxy-HW condition, relative to the other conditions, but only on Day 1. Among those receiving treatment, positive and negative expectations, and treatment-related worry, did not vary by interaction-style, while psychological well-being and side effect reporting were inversely associated at the level of the individual. Results have important implications for practice, suggesting poorer quality interactions may not only reduce beneficial health outcomes but also exacerbate those that are maladaptive.