The trade of wild epiphytic orchids has put many populations at risk of extinction; however, little is known about the impact of repeated harvesting of individuals on population dynamics. A population viability analysis (PVA) was developed to simulate the impact of harvesting on the risk of extinction of a population of an epiphytic orchid that is sold in Mexican markets. In January 2004, 388 individuals of Guarianthe aurantiaca were labeled and measured. The same individuals were measured again in December 2005 and December 2006. Capsule production and mortality were also recorded. Deterministic (d) and stochastic (s) matrix models were developed to estimate the population growth rate (λ). A PVA was developed to evaluate the probability of extinction over a 100-year period, given a threshold of 5% of the initial population and different harvest intensities of G. aurantiaca. The λd values were statistically indistinguishable from unity (0.989 ± 0.103 for 2004–2005; 0.990 ± 0.087 for 2005–2006); the λs values were below unity (0.988 ± 0.001). The reproductive stages had the highest elasticity values, whereas the persistence of individuals in their category was the demographic process with the highest elasticity. A life table response experiment showed that the difference in the λd values was accounted for by the positive contribution of the stasis of individuals in category j and the retrogression of individuals in category r1 to category j. The extinction risk was 100% when more than 5% of the reproductive individuals were harvested. The results suggest that this G. aurantiaca population is in a precarious equilibrium and harvesting should be controlled and restricted to immature individuals.