Lattice structures for implants can be printed using metal three-dimensional (3D)-printing and used as a porous microstructures to enhance bone ingrowth as orthopedic implants. However, designs and 3D-printed products can vary. Thus, we aimed to investigate whether targeted pores can be consistently obtained despite printing errors. The cube-shaped specimen was printed with one side 15 mm long and a full lattice with a dode-thin structure of 1.15, 1.5, and 2.0 mm made using selective laser melting. Beam compensation was applied, increasing it until the vector was lost. For each specimen, the actual unit size and strut thickness were measured 50 times. Pore size was calculated from unit size and strut thickness, and porosity was determined from the specimen’s weight. The actual average pore sizes for 1.15, 1.5, and 2.0 mm outputs were 257.9, 406.2, and 633.6 μm, and volume porosity was 62, 70, and 80%, respectively. No strut breakage or gross deformation was observed in any 3D-printed specimens, and the pores were uniformly fabricated with < 10% standard deviation. The actual micrometer-scaled printed structures were significantly different to the design, but this error was not random. Although the accuracy was low, precision was high for pore cells, so reproducibility was confirmed.