An apparatus for multiaxial loading of microtubes with an outside diameter of less than a few millimeters is described. The apparatus is capable of applying axial-force and internal-pressure in a fully-coupled way, so that biaxial principal stresses along either proportional or non-proportional loading paths can be applied to the microtubes. The requirements of the apparatus for this fully-coupled loading and the resulting functional layout are described first. The hardware that was selected is then detailed. The maximum force and pressure capacities are 2000 N and 1034 bar, while the maximum available stroke and pressurized fluid volume are 50 mm and 68 ml, respectively. Since the apparatus is required to perform dynamic loading as well, with the stresses varying on the order of 1 Hz, the dynamic behavior of the pump used in the apparatus is established experimentally. It was determined that the dynamic response depends not only on the amplitude and frequency of the desired loading, but also, partially because of the minute dimensions of the apparatus, on the volume of fluid in the system and the compressibility of the fluid used for the inflation. The capabilities of the apparatus are demonstrated by proportional and non-proportional biaxial experiments on 304L stainless steel microtubes of 2.38 mm outside diameter and 0.15 mm nominal thickness. Two different strategies for performing these experiments, i.e., force- & volume-control and displacement- & pressure-control were implemented and are compared to each other. The former leads to a more stable system and thus is the preferred mode of operation of the apparatus, while the latter can allow better tracking of the microtube response past the observed limit-load instability. In the proportional loading experiments, the nominal stress paths are shown to be exactly linear, as desired. Although not pursued here, such experiments can be used to establish the plastic anisotropy of the microtube material. Two types of non-proportional experiments are performed: in both cases the microtube is loaded uniaxially (axial or hoop direction); then, by keeping that stress constant, the other principal stress (i.e., hoop or axial) is increased until failure. Digital Image Correlation is used to measure the full strain fields during the testing, establishing the strains at the limit of uniform deformation. Furthermore, comparison of the strain paths induced during the proportional and non-proportional experiments established the path-dependency of the failure strains. These successful experiments demonstrate that any complex loading path inside (or adjacent to) the first quadrant of the plane-stress space can be implemented with this apparatus.