Translocation gains importance as a conservation approach of choice for imperiled plant species whose habitat is rapidly disappearing. To be successful, translocation must be done into a location that has optimal conditions for the species, but determining these conditions can be difficult for species with poorly known historic distribution. Unfortunately, most endangered plant species, including those requiring immediate action, have poorly documented historic ranges. In this study, we show, using the threatened rhizomatous Iris atrofusca, that the determination of suitable locations should be done experimentally over multiple locations, and the evidence for the success of translocation comes not just from the establishment of the planted individuals or their reproduction, but, more importantly, from the appearance, establishment and reproduction of recruits. Although 12 years after introduction, 2 of the 22 populations went extinct and the number of survived plants was below 10% in 12 populations, 4 established recruits were detected outside the planted areas in that year and another 48 individuals 2 years later. We conclude that locations in the immediate vicinity of recruits should be used for actual translocation. We also report, after analyzing the effect of four environmental variables on I. atrofusca performance across experimental microsites, the most favorable set of topographic and soil parameters for this species, namely deep soil having few rock fragments and minor slope steepness. Microsites with these parameters should be the preferred locations for I. atrofusca introduction. Overall, our study demonstrates the importance of (1) experiments preceding actual translocation and (2) long-term monitoring for translocation projects.