The use of small specimen geometries in asphalt mixture performance testing to enable the testing of as-built pavement layers has been gaining attention in recent years. Small specimens could also improve the testing efficiency of laboratory-fabricated specimens by allowing the extraction of multiple test specimens per gyratory-compacted sample. Rigorous assessment of the small specimen geometries is required before the use of such geometries is standardized. In this study, small specimens were evaluated for dynamic modulus and simplified viscoelastic continuum damage fatigue. Three specimen geometries (100-mm- and 38-mm-diameter cylindrical specimens and 25- × 50-mm prismatic specimens) were compared by using five mixtures with a nominal maximum aggregate size (NMAS) ranging from 9.5 to 25.0 mm. The results show that the dynamic modulus and phase angle master curves agreed at low and intermediate temperatures, regardless of the NMAS values of the mixture. At the high temperature, the small specimen dynamic modulus values were slightly higher and the phase angle values were slightly lower than those of the large specimens. The specimen-to-specimen variability for the large and small specimens was comparable. The fatigue test results for the mixtures evaluated were comparable, except for the 25-mm mixture, which proved problematic in the testing of both small and large specimens. Pavement performance was predicted by the layered viscoelastic analysis for critical distresses program by using the test results for the small and large specimens. These results suggest that specimen geometry had a minimal effect on pavement fatigue damage predictions, which indicates promise for the use of small specimen geometries in practice.