The San Francisco Emergency Medical Services study25 was a randomized, double-blind trial to evaluate IV benzodiazepine administration by paramedics for the treatment of out-of-hospital patients with status epilepticus.25 In this study, 205 patients were randomized to IV diazepam (5 mg), lorazepam (2 mg), or placebo. An identical second injection was administered if needed. Status epilepticus had terminated at arrival in the emergency department in 59.1% of the patients treated with lorazepam, in 42.6% of the patients treated with diazepam, and in 21.1% of patients treated with placebo (lorazepam vs diazepam: odds ratio, 1.9; 95% confidence interval [CI], 0.9 to 4.3). The duration of the status epilepticus was shorter in the lorazepam group compared to the diazepam group (adjusted relative hazard, 0.65; 95% CI, 0.36 to 1.17). These data are supported by a double-blind study reported by Leppik et al60 in 1983 in which 78 patients with status epilepticus were randomized to receive one or two doses of either lorazepam (4 mg) or diazepam (10 mg). Seizures were controlled in 89% of the episodes treated with lorazepam and in 76% of those treated with diazepam. Although the dosages of lorazepam and diazepam differed in these three studies and phenytoin was added to diazepam in the VA study,25 the summed data indicate that lorazepam is significantly more effective in terminating seizures than is diazepam (odds ratio, 1.74; 95% CI, 1.14 to 2.64; p = 0.01). Furthermore, the pharmacokinetic properties of lorazepam favor its use over that of diazepam. The anticonvulsant effect of a single dose of diazepam is very brief (20 min), whereas that of lorazepam is much longer (> 6 h), and the risk of respiratory depression may be greater with diazepam.61 Although diazepam has a much longer elimination half-life, due to its high lipid solubility it is rapidly redistributed from the brain to the peripheral fat stores, accounting for its shorter antiseizure activity.