Pharmacological manipulation of memory reconsolidation opens up promising new avenues for anxiety disorder treatment. However, few studies have directly investigated reconsolidation-based approaches in subclinical or clinical populations, leaving optimal means of fear memory reactivation unknown. We conducted a systematic pilot study to assess whether a reconsolidation-based treatment could tackle public speaking anxiety in a subclinical sample (N = 60). As lab studies indicate that the duration of reactivation may be important for inducing reconsolidation, we investigated several speech lengths to help inform further translational efforts. Participants underwent a stress-inducing speech task composed of 3-min preparation, and from 0 to 9 min of public speaking, in 1-min increments. They then received either 40 mg of propranolol (n = 40) or placebo (n = 20), double-blind, allocated 4:2 for each speech duration. Participants performed a second speech 1 week post treatment, and were followed up with questionnaires 1- and 3 months later. Both self-reported speech distress and questionnaire measures of public speaking anxiety showed clear reductions following treatment. However, propranolol did not reliably outperform placebo, regardless of speech duration at treatment. Physiological responses (heart rate and salivary cortisol) to the public speaking task remained stable from treatment to test. These findings highlight the challenges facing the translation of laboratory research on memory reconsolidation into clinical interventions. Lack of explicit controls for factors beyond duration, such as 'prediction error', could explain these null findings, but positive results in clinical interventions are needed to demonstrate that taking such factors into account can deliver the promises of reconsolidation-based therapy.