This paper analyses decisions regarding smoking and drinking for a sample of Irish women. Double-hurdle models are estimated to determine whether decisions to smoke/drink are made independently of how much to smoke/drink. Given the potential complementarities between smoking and drinking a model which allows for the joint determination of smoking and drinking is also estimated. The paper finds that decisions to smoke/drink and how much to smoke/drink are not independent and that decisions to smoke/drink are not made independently of each other.