Blended learning, mixing online and face-to-face learning modes, has been increasingly adopted in training with the expectation that it will help solve transfer problems and will lend other educational benefits. While there is a set of knowledge about design for facilitating transfer, it has not been investigated in a blended environment context. In response, this study investigated training design related factors that facilitated and hindered transfer in a blended-training context. The case program was a 6-week management leadership development program offered by a corporate university in Korea. Both facilitating and hindering design factors at each learning node were investigated by conducting a literature review, document analyses, focus group interviews, one-on-one interviews and an online survey. In addition to descriptive analysis, a smaller number of factors were drawn as salient design factors having predictability for transfer by using stepwise multiple regression. For the online module, the facilitating factors in "demonstration" and "activation" were best predictors for transfer. As a predictor among the hindering factors in the online module, the factors in "demonstration" were yielded. For the off-line module, the factors in "application" were drawn as predictor factors. However, no single hindering variable resulted from the multiple regressions related to this module. By integrating the descriptive and predictive statistical analysis with qualitative data analysis, specific design guidelines for blended training to ensure transfer were presented.