Objectives: This study compared the flow and filling of several retrograde filling materials using new different test models. Materials and Methods: Glass plates were manufactured with a central cavity and 4 grooves in the horizontal and vertical directions. Grooves with the dimensions used in the previous study (1 × 1 × 2 mm; length, width, and height respectively) were compared with grooves measuring 1 × 1 × 1 and 1 × 2 × 1 mm. Biodentine, intermediate restorative material (IRM), and mineral trioxide aggregate (MTA) were evaluated. Each material was placed in the central cavity, and then another glass plate and a metal weight were placed over the cement. The glass plate/material set was scanned using micro-computed tomography. Flow was calculated by linear measurements in the grooves. Central filling was calculated in the central cavity (mm3) and lateral filling was measured up to 2 mm from the central cavity. Results: Biodentine presented the least flow and better filling than IRM when evaluated in the 1 × 1 × 2 model. In a comparison of the test models, MTA had the most flow in the 1 × 1 × 2 model. All materials had lower lateral filling when the 1 × 1 × 2 model was used. Conclusions Flow and filling were affected by the size of the test models. Higher grooves and materials with greater flow resulted in lower filling capacity. The test model measuring 1 × 1 × 2 mm showed a better ability to differentiate among the materials.