In vitro methods for propagation and preservation are critical when seed banking is not an option for the ex situ conservation of an endangered species. Such methods rely on the ability to maintain normal growth for the ultimate goal of producing plants. Previous studies revealed that the threatened Cycladenia humilis var. jonesii can exhibit extreme hyperhydricity when cultured in vitro, but that this phenotype can be largely normalized by venting of the cultures. A design of experiments (DOE) approach was used to investigate factors contributing to hyperhydricity in this species. Using a 6-factor fractional factorial design, the effects of nitrogen, CaCl2, BAP, gelling agent, and the presence of venting on the phenotype were tested. Venting was the strongest driver decreasing hyperhydricity, but gelling agent and other factors had significant effects, depending on the response measured. A revised medium was then designed and compared with the original maintenance medium through 2–3 subcultures. Venting was again the strongest driver decreasing hyperhydricity, but the revised medium also significantly affected some responses, even in the absence of venting. The revised medium was not optimal, as shoots became progressively smaller with each transfer, but despite their size, hyperhydricity decreased. These results demonstrate the usefulness of DOE for simultaneous testing of multiple factors, but also the need for follow-up optimization experiments. DOE can be particularly useful in working with rare species. By making efficient use of limited resources, it can facilitate the development of in vitro propagation protocols for such species and help ensure their conservation. [ABSTRACT FROM AUTHOR]