Direct ink writing, or robocasting, is an extrusion additive manufacturing technique for the fabrication of complex ceramic parts. Pores are common defects in post-sintered robocast parts that strongly influence the performance by changing the density, the transport properties, and the mechanical strength. In this work, the porosity (volumetric fraction, size distribution, geometry and topological distribution) of monolithic and 3D-lattice specimens made of hydroxyapatite was comprehensively characterized at multiple length scales through the six most widely used experimental methods for the study of porous materials. These two types of samples embrace the two most common types of additive manufactured ceramics and allowed the study of materials with pores in the submicron scale, as well as materials with a bimodal pore size distribution at significantly different length scales. Detected pores were divided into (1) engineered porosity set by the structural design, and (2) hollow defects, including intergranular porosity, trapped-air pores, cracks, and cavities, that overlapped at different length scales with the engineered porosity. The origin and mechanisms of formation of hollow defects are discussed, providing guidelines to avoid them. The experimental methods that allow discerning between pores and hollow defects are highlighted, and their advantages and drawbacks are discussed. This work might serve as a guide for the selection of the proper combination of methods for the pore evaluation of similar additive manufactured parts. • Hollow defects in robocast ceramic objects were characterized from the nano to the over micrometric scale. • Image analysis, He pycnometry, N 2 adsorption, MIP, µ-CT, and water saturation were used for characterization. • The types of hollow defects include intergranular porosity, trapped-air pores, cracks, and cavities. • The origin and mechanism of formation of hollow defects is discussed, providing general guidelines to avoid them. • Discussion of the advantages and limits of the methods for the evaluation of materials with a bimodal pore size distribution.