A household survey of Tsimane' Indians of the Bolivian rain forest is used to examine the effect of different types of misfortunes (e.g., illness, deaths, evacuations, crop loss) at the level of the household, village, and region on different types of coping mechanisms. Forest clearing and advances on salaries seem to be more important than savings, reciprocity, borrowing, or credit in coping with mishaps. The statistical evidence suggests that Tsimane' remain poorly insured against unforeseen misfortunes and that inter- or intravillage reciprocity is not operative in times of need. (Misfortunes, Tsimane', food security, insurance, reciprocity, credit, rain forest, Bolivia) Since the early 1950s scholars have been saying that rain forests provide security to rural people, particularly when they face random misfortunes such as crop losses (e.g., Falconer and Koppell 1990:98). During mishaps dwellers of the forest are said to increase foraging and other forms of dependence on the forest. The point has been made chiefly through qualitative case studies from Africa, though Ogle (1996) and others (e.g., Scoones et al. 1992) have found similar evidence in Asia and Latin America. This article takes a first step in examining the insurance value of forests to indigenous people by asking the broader question of how relatively isolated rural household in the rain forest protect themselves against different types of unanticipated misfortunes, not just against crop losses. A recently completed household survey of the Tsimane' Indians of the Bolivian rain forest tested the extent to which different regional, village, household, and personal misfortunes, such as deaths in the family, illness, medical emergencies, and crop losses, affect a household's decisions to: 1) clear old-growth or secondary-growth rain forest; 2) seek wage labor or advances on salaries; 3) use credit; 4) borrow seeds from neighbors; 5) deplete their own savings in animals; and 6) rely on intra- or intervillage reciprocity. The study is novel in two ways. First, through a multivariate framework the received wisdom that forest-dwellers increase their reliance on the forest in times of need is tested. The intuition so far has not been put to a rigorous quantitative test. Second, this article contributes to the growing literature on rural insurance mechanisms (Alderman and Paxson 1992; Morduch 1995; Townsend 1995) by examining how people living in relative autarky cope with misfortunes; most previous research has come from peasants with stronger links to the market. Many researchers have said that rural people increase their dependence on the forest when they face unanticipated misfortunes (Gunatilake et al. 1993; Hecht et al. 1988; Towson 1994; Scoones et al. 1992; Falconer and Koppell 1990; Falconer and Arnold 1989; Falconer 1992; Campbell-Platt 1980). The forest offers two forms of insurance to smallholders: relatively liquid assets (such as game, honey, tubers, seeds, fruits, construction materials, and medicines) and relatively nonliquid assets (such as the nutrients locked in the soil and phytomass). The former can be harvested quickly during crisis; the latter can be diverted to annual or perennial crops as a longer-term form of insurance. How the forest is used as a safety net deserves closer empirical scrutiny for at least three reasons. First, most case studies of people's growing reliance on the forest in times of need have been qualitative (Ogle 1996). The idea so far has not been tested with multivariate techniques. Households that suffer misfortunes may need to increase their dependence on the forest only if they lack cheaper forms of self-insurance. Once researchers control for reciprocity, credit, or for remittances from urban kin, they may find that misfortunes do not drive dependence on the forest. Second, though crop losses may increase reliance on nonliquid forest assets, other types of misfortunes (e. …