The existing literature on precautionary saving is based almost entirely on the assumption that the household acts as if it consisted of a single individual decision-taker. In reality saving decisions are typically taken by two-person households. This paper examines the implications of this observation for the existence of precautionary saving, and shows that the assumption that the individual utility functions satisfy the conditions for precautionary saving to exist can imply that the household exhibits precautionary saving, but only under strong assumptions on the type of risk change being considered. (JEL D14, D81)