Background: This study examined whether biological mechanisms linking dementia caregiving with an increased risk of coronary heart disease can be modified by psychosocial interventions and which caregivers might benefit the most from an intervention. Methods: Spousal dementia caregivers were randomized to 12-week treatment with either a behavioral activation intervention (ie, Pleasant Events Program [PEP]; n = 60), or an active control Information and Support (IS; n = 63) condition. Indicators of caregiving stress were assessed pretreatment and circulating cardiovascular biomarkers were measured pre- and posttreatment. Results: There were no significant changes in biomarker levels from pre- to posttreatment both by treatment condition and across all caregivers. Regardless of the treatment condition, exploratory regression analysis revealed that caregivers were more likely to show significant decreases in C-reactive protein (CRP) and D-dimer when their spouse had severe functional impairment; in interleukin (IL)-6 and CRP when they had greater distress due to care recipient's problem behaviors; in tumor necrosis factor (TNF)-α when they had higher levels of negative affect; and in IL-6, CRP, TNF-α, and D-dimer when they had higher personal mastery. Within the PEP group, caregivers with higher negative affect and those with higher positive affect were more likely to show a reduction in von Willebrand factor and D-dimer, respectively. Within the IS group, caregivers whose spouse had severe functional impairment were more likely to show a decrease in IL-6. Conclusions: Unlike the average caregiver, caregivers high in burden/distress and resources might benefit from psychosocial interventions to improve cardiovascular risk, although these observations need confirmation.