The article attempts to estimate the efficiency of doctoral studies in 58 public higher education institutions in Poland from 2010 to 2015. The authors employ a hybrid model combining radial and non-radial approaches as part of the non-parametric Data Envelopment Analysis (DEA) method. An assessment of both the direct and indirect effectiveness of doctoral studies has been made: from the beginning of studies to obtaining a doctoral degree, and from the start of studies to the launch of a doctoral programme respectively. In the case of both efficiency and effectiveness measurements, two empirical models with different research periods were used. In the first empirical model, the efficiency and effectiveness of studies were estimated at their nominal duration (four years), while in the second model the research period was extended by one year. Only four universities proved to be efficient in the first model and five in the second, while none turned out to be fully effective. Probably the higher level of efficiency is due to high competition between universities in the largest urban conurbations. On the other hand, the effectiveness of doctoral studies is to an extent correlated with the type of higher education institution (technical, pedagogical, etc.). Extending the research period by one year has a greater impact on the efficiency rather than effectiveness of doctoral studies.