A new study may help drinkers pick their poison. In a head-to-head comparison, bourbon gave drinkers a more severe hangover than vodka, report Damaris Rohsenow of Brown University and colleagues in an upcoming issue of Alcoholism: Clinical and Experimental Research.

But vodka drinkers aren’t off the hook: Drinkers’ sleep suffered equally with both drinks, as did their performance on tasks requiring attention and quick responses. Understanding the lingering effects of alcohol after a night of heavy drinking is important for people who engage in safety-sensitive tasks, such as driving, while hung over Rohsenow says.

The researchers recruited 95 healthy young adults, ages 21 to 33, and gave them caffeine-free cola mixed with bourbon, vodka or tonic water. The drinking ended when participants’ breath-alcohol concentrations hit an average of 0.11, well over the legal intoxication limit. Participants were then hooked up to sleep monitors, which record brain activity, and allowed to sleep it off. At 7 a.m. the next day, the researchers roused the subjects from bed (a wake-up that did not include coffee or aspirin) and asked them to rate the severity of their hangovers.

Overall, bourbon drinkers reported feeling worse than vodka drinkers, rating higher on scales that measure the severity of hangover malaise, including headache, nausea, loss of appetite and thirst. It should come as no surprise that alcohol drinkers said they felt much worse than those who had drunk only tonic water.

One reason for the different effects of vodka and bourbon, Rohsenow says, could be that bourbon contains 37 times more toxic compounds than vodka does, including nasty organic molecules such as acetone, acetaldehyde, tannins and furfural. A good rule of thumb for liquors, she says, is that the clearer they are, the less of these substances they contain.

Both the bourbon drinkers and vodka drinkers slept poorly compared to the nondrinkers, the team found. The next morning, when the participants performed cognitive tests that required attention and quick reaction times, the drinkers performed worse than the nondrinkers, but the type of alcohol had no effect on performance. Both groups of drinkers were impaired equally.