The oxidative stress theory of aging proposes that accumulation of oxidative damage is the main proximate cause of aging and that lifespan is determined by the rate at which this damage occurs. Two predictions from this theory are that long-lived organisms produce fewer ROS or have increased antioxidant production. Based in these predictions, molecular mechanisms to promote longevity could include either changes in the regulation of mitochondrial genes that affect ROS production or elevated expression of antioxidant genes. We explored these possibilities in the honey bee, a good model for the study of aging because it has a caste system in which the same genome produces both a long-lived queen and a short-lived worker. We measured mRNA levels for genes encoding eight of the most prominent antioxidant enzymes and five mitochondrial proteins involved in respiration. The expression of antioxidant genes generally decreased with age in queens, but not in workers. Expression of most mitochondrial genes, in particular CytC, was higher in young queens, but these genes showed a faster age-related decline relative to workers. One exception to this trend was COX-I in thorax. This resulted in higher COX-I/CytC ratios in old queens compared to old workers, which suggests caste-specific differences in mitochondrial function that might be related to the caste-specific differences in longevity. Queen honey bee longevity appears to have evolved via mechanisms other than increased antioxidant gene expression.