Although doctoral mentors recognize the benefits of providing quality advisement and close guidance, those of sharing project management responsibilities with mentees are still not well recognized. We observed that mentees, who have the opportunity to co-manage projects, generate more written output. Here we examine the link between research productivity, doctoral mentoring practices (DMP), and doctoral research experiences (DRE) of mentees in programs in the non-West. Inspired by previous findings that early career productivity is a strong predictor of later productivity, we examine the research productivity of 210 molecular biology doctoral students in selected programs in Japan, Singapore, and Taiwan. Using principal component (PC) analysis, we derive two sets of PCs: one set from 15 DMP and another set from 16 DRE items. We model research productivity using Poisson and negative-binomial regression models with these sets as predictors. Our findings suggest a need to re-think extant practices and to allocate resources toward professional career development in training future scientists. We contend that doctoral science training must not only be an occasion for future scientists to learn scientific and technical skills, but it must also be the opportunity to experience, to acquire, and to hone research management skills.