Following the approach of the National Primary Laboratory of the UK (NPL) for the calibration of radionuclide calibrators, but using a commercially available instrument with no data available in the literature, the radionuclide calibrator response was investigated as a function of different measurement geometries at the "Regina Elena" National Cancer Institute (IRE) in Rome. Working with Italian National Metrology Institute for ionising radiation quantities (ENEA-INMRI), specific calibration factors with traceability to national primary standards were determined for different types of glass vials, solid capsules and plastic syringes, investigating three radionuclides with different energy spectra (Tc-99m, In-111, I-131). For each kind of syringe, calibration correction factors for different filling volumes were calculated. For Tc-99m and I-131 the difference between measured and true activity was in the range 2-7%, depending on measurement geometry. For In-111 a large percentage deviation from the true activity value was found in each geometry considered, reaching 35%. The magnitude of this difference is particularly dependent on the energies of the emitted photons.