Additive manufacturing and 3D printing is particularly useful in the production of phantoms for medical imaging applications including determination and optimization of (diagnostic) image quality and dosimetry. Additive manufacturing allows the leap from simple slab and stylized to (pseudo)-anthropomorphic phantoms. This necessitates the use of materials with x-ray attenuation as close as possible to that of the tissues or organs mimicked. X-ray attenuation properties including their energy dependence were determined for 35 printing materials comprising photocured resins and thermoplastic polymers. Prior to measuring x-ray attenuation in CT from 70 to 140 kVp, printing parameters were thoroughly optimized to ensure maximum density avoiding too low attenuation due to microscopic or macroscopic voids. These optimized parameters are made available. CT scanning was performed in a water filled phantom to guarantee defined scan conditions and accurate HU value determination. The spectrum of HU values covered by polymers printed using fused deposition modeling reached from -258 to +1,063 at 120 kVp (-197 to +1,804 at 70 kVp, to -266 to +985 at 140 kVp, respectively). Photocured resins covered 43 to 175 HU at 120 kVp (16-156 at 70, and 57-178 at 140 kVp). At 120 kVp, ASA mimics water almost perfectly (+2 HU). HIPS (-40 HU) is found close to adipose tissue. In all photocurable resins, and 17 printing filaments HU values decreased with increasing beam hardness contrary to soft tissues except adipose tissue making it difficult to mimic water or average soft tissue in phantoms correctly over a range of energies with one single printing material. Filled filaments provided both, the HU range, and an appropriate energy dependence mimicking bone tissues. A filled material with almost constant HU values was identified potentially allowing mimicking soft tissues by reducing density using controlled under-filling. The measurements performed in this study can be used to design phantoms with a wide range of x-ray contrasts, and energy dependence of these contrasts by combining appropriate materials. Data provided on the energy dependence can also be used to correct contrast or contrast to noise ratios from phantom measurements to real tissue contrasts or CNRs.