Gabapentin has been reported to be effective for essential tremor (ET) based on open-label trials. We studied gabapentin (1800 mg/day) and placebo in a double-blind crossover design in 20 ET patients. Eighteen patients completed the study and two patients dropped out as a result of adverse effects which resolved when the medication was discontinued. Tremor was assessed at baseline and after 2 weeks of gabapentin and placebo treatment. One patient was mildly improved and another was moderately improved with placebo. Similarly, one patient reported mild improvement and another patient had marked improvement with gabapentin. All the remaining patients either reported no change or were worse with both treatment arms. There was no significant difference for total tremor score, hand tremor score, handwriting scores, or pouring scores. Sickness Impact Profile scores were no different between placebo and gabapentin. Our results suggest that as an adjuvant therapy in ET, gabapentin has limited benefit.