Vegetative dormancy is a life history phenomenon in which herbaceous perennial plants fail to produce fully aboveground sprouts. Its evolutionary ecology has been hypothesized to be potentially maladaptive in some cases, and potentially adaptive in others. In particular, the adaptive hypothesis is based on the possibility that dormancy is a means of weathering environmental stochasticity. Here, we present a study on two dormancy-prone perennial species, Cypripedium calceolus and Cephalanthera longifolia, studied in five population in Estonia. Approx. 10-20 plants at each site were defoliated, a further 10-20 were shaded, and a further 10-20 were used as controls, with treatment continuing for two years. The study was begun in May 2002, and monitoring continued every year through 2007. We analyzed the data using logistic regression of observed lifespan as a function of treatment and observed dormancy events, and fixed-effect life table response experiments (LTRE) to test for impacts on population growth rate and fitness. Both dormancy and mortality responded to treatment, and plants that went dormant more often had lower mortality by 2007. Dormancy had a significantly positive impact on population growth rate, with increased dormancy associated with unusually high increases in growth rate. Hence, vegetative dormancy appears to act adaptively over the long-term.