Background There are no medications approved for as‐needed use for feared situations for individuals with social anxiety disorder (SAD). In the present study, intranasal PH94B was provided for use as needed during stressful events. Methods Twenty‐two subjects were randomized (double‐blind) to 2 weeks of treatment with intranasal PH94B or placebo. Following self‐administration of medication prior to a feared event, peak levels of anxiety were recorded using the Subjective Units of Distress Scale (SUDS). After 2 weeks, subjects were crossed over to the opposite treatment for 2 weeks. Average peak SUDS during treatment with PH94B and placebo were compared using a paired t‐test. Results Significant differences in favor of PH94B were found on the primary outcome measure: mean peak SUDS change from baseline for all subjects receiving PH94B was 15.6 points versus 8.3 points for placebo (paired t = 3.09, P = .006, effect size of .658). PH94B showed less superiority over placebo when placebo was given second rather than first, likely due to a carryover effect. Looking between groups at just the first 2 weeks of treatment, PH94B also showed trend superiority to placebo on the Liebowitz Social Anxiety Scale (LSAS) (P = .07) and a significant difference on the Patient Global Impression of Change (P = .024) and the LSAS Avoidance subtotal (P = .02). Conclusions While further study is needed, these results, combined with earlier findings, suggest that PH94B could represent a useful as‐needed treatment for SAD, and continue to validate the nasal chemosensory system as a novel mechanism for medication delivery.