Despite frequent benzodiazepine use in anxiety disorders, the trajectory and magnitude of benzodiazepine response and the effects of benzodiazepine potency, lipophilicity, and dose on improvement are unknown. We performed a meta-analysis using weekly symptom severity data from randomized, parallel group, placebo-controlled trials of benzodiazepines in adults with anxiety disorders. Response was modeled for the standardized change in continuous measures of anxiety using a Bayesian hierarchical model. Change in anxiety was evaluated as a function of medication, disorder, time, potency, lipophilicity, and standardized dose and compared among benzodiazepines. Data from 65 trials (73 arms, 7 medications, 7110 patients) were included. In the logarithmic model of response, treatment effects emerged within 1 week of beginning treatment (standardized benzodiazepine-placebo difference = -0.235 ± 0.024, CrI: -0.283 to -0.186, P < .001) and placebo response plateaued at week 4. Doses <6 mg per day (lorazepam equivalents) produced faster and larger improvement than higher doses (P = .039 for low vs medium dose and P = .005 for high vs medium dose) and less lipophilic benzodiazepines (beta = 0.028 ± 0.013, P = .030) produced a greater response over time. Relative to the reference benzodiazepine (lorazepam), clonazepam (beta = -0.217 ± 0.95, P = .021) had a greater trajectory/magnitude of response (other specific benzodiazepines did not statistically differ from lorazepam). In adults with anxiety disorders, benzodiazepine-related improvement emerges early, and the trajectory and magnitude of improvement is related to dose and lipophilicity. Lower doses and less lipophilic benzodiazepines produce greater improvement.