Smiling has been previously shown to improve stress responses. We replicated and expanded this work by testing whether smiling helps with a potent real-world stressor: a vaccination-like needle injection. We also extended past research by examining grimacing, a facial expression known to naturally occur during stress and pain and one that shares some of the same facial action units as smiling. Participants (n = 231; [M]age = 19.2) were randomized to hold either a Duchenne smile, a non-Duchenne smile, a grimace, or a neutral expression while receiving a 25-gauge needle injection of saline solution. Expression was covertly manipulated via cover story and chopstick placement in the mouth. Heart rate (HR) and electrodermal activity (EDA) were collected continuously alongside self-reports of pain, emotion, and distress. Repeated-measures analysis of variance (ANOVA) indicated a between-subjects effect of facial condition on self-reported pain as well as a Condition × Time effect. Probing each time point revealed the strongest effect to be at needle injection, where the Duchenne smile and grimace groups reported approximately 40% less needle pain versus the neutral group. Repeated-measures ANOVAs also revealed differences between conditions for both HR and EDA. In post hoc analyses, only the Duchenne smile group exhibited significantly lower HR than neutral, with marginal Duchenne benefits found for EDA. Together, these findings indicate that both smiling and grimacing can improve subjective needle pain experiences, but Duchenne smiling may be better suited for blunting the stress-induced physiological responses of the body versus other facial expressions.