This paper investigates which attributes of a Carnegie PhD-level institution predict the share and number of its undergraduate humanities BA recipients that will go on to earn a humanities PhD. We use restricted-access individual-level Survey of Earned Doctorates data from the National Science Foundation (NSF) to determine both where and when PhD recipients received their BA. We use a truncation-correction methodology to account for problems inherent with studying PhD recipients, who often will receive their PhD after the data end. Using OLS, negative binomial regression, and an analysis similar to that of a prior, related paper, we find robust relationships between PhD production and student test scores, instructional expenditures per student, and the number of highly-ranked humanities PhD programs an institution has.