This study examines the possibility that volunteer support can influence how long terminally ill patients survive. Hospice patient files (N = 290) were coded for marital status and volunteer support condition, respectively, the latter on.the basis of whether visits from volunteers were requested and received (n = 94), requested but not received (n = 28), or neither requested nor received (n = 168). Baseline health, disease type, and demographic dimensions were comparable across support conditions. Results indicated that when a baseline health status effect was controlled for (p < .0002), patients in the volunteer support condition survived significantly longer than did patients in either unvisited condition (p < .0001). Neither marital status nor gender independently predicted survival time.