For more than 20 years, a plethora of research has been conducted on the discounting of delayed rewards. In contrast, there has been relatively little research on the discounting of delayed aversive outcomes (e.g., monetary losses). The present study examined the discounting of delayed hypothetical gains and losses by 55 college undergraduates at 2 monetary amounts, $1,000 and $25,000. A simple hyperbola often provided very poor fits to the data. In contrast, a hyperboloid generally provided much better fits to the discounting of all outcomes, and its exponent was frequently less than unity. Although the rate of discounting was greater for the small gain than for the large gain (i.e., a magnitude effect), the rate of discounting for the small and large loss was not significantly different, which led to a gain-loss asymmetry for the small amount (i.e., small gains discounted more than small losses), but not for the large amount (i.e., large gains discounted similarly to large losses). Collectively, these findings suggest that similar but separate processes underlie the discounting of delayed gains and losses.