The search for novel anticonvulsants requires appropriate model systems in which to test hypotheses through focused compound screening or genetic manipulation, or conduct black box screening of large numbers of compounds or potential genetic modifiers. Many models are currently in existence that subserve particular roles in achieving these aims, but all have their limitations. Zebrafish have been suggested as an additional model of epilepsy, but their optimum role is unclear. They are more amenable to high throughput analysis, but are more genetically removed from humans than rodents. We therefore sought to develop assay methodology applicable to medium/high throughput screening using an automated tracking system to measure the amount of movement induced by exposure to the proconvulsant, pentylene tetrazole (PTZ). We then used this system to explore how many known anti-epileptic drugs (AEDs) would be detected when running such a screen. We were able to detect suppression of PTZ-induced excessive movements with 13 out of 14 standard AEDs. A parallel sedation and toxicity screen suggested these effects were due to direct anti-epileptic effect, although non-specific effects cannot be fully excluded. These results suggest zebrafish may be a useful high throughput primary screen to pick up potential novel AEDs.