For pediatric femoral shaft fractures, elastic stable intramedullary nailing (ESIN) is an accepted method of treatment. But problems regarding stability with shortening or axial deviation are well known in complex fracture types and heavier children. Biomechanical in vitro testing was performed to determine whether two modified osteosyntheses with an additional tension screw fixation or screw fixation alone without nails could significantly improve the stability in comparison to classical ESIN. A total of 24 synthetic adolescent-sized femoral bone models (Sawbones, 4th generation; Vashon, Washington, United States) with an identical spiral fracture (length 100 mm) were used. All grafts underwent retrograde fixation with two C-shaped steel nails (2C). Of the 24, 8 osteosyntheses were supported by one additional tension screw (2C1S) and another 8 by two screws (2S) in which the intramedullary nails were removed before testing. Each configuration underwent biomechanical testing in 4-point bending, external rotation (ER) and internal rotation (IR). Furthermore, the modifications were tested in axial physiological 9 degrees position for shifting and dynamic compression as well as dynamic load. Both screw configurations (2C1S and 2S) demonstrated a significantly higher stability in comparison to the 2C configuration in 4-point bending (anterior-posterior, 0.95 Nm/mm [2C] < 8.41 Nm/mm [2C1S] and 15.12 Nm/mm [2S]; posterior-anterior, 8.55 Nm/mm [2C] < 12.65 Nm/mm [2C1S] and 17.54 Nm/mm [2S]; latero-medial, 1.17 Nm/mm [2C] < 5.53 Nm/mm [2C1S] and 9.15 Nm/mm [2S]; medio-lateral, 1.74 Nm/mm [2C] < 9.69 Nm/mm [2C1S] and 12.20 Nm [2S]; all p < 0.001) and during torsion (ER, 0.61 Nm/degree [2C] < 4.10 Nm/degree [2C1S] and 9.29 Nm/degree [2S]; IR, 0.18 Nm/degree [2C] < 6.17 Nm/degree [2C1S] and 10.61 Nm/degree [2S]; all p < 0.001]. The shifting in compression in 9 degrees position was only slightly influenced. The comparison of 2S versus 2C1S showed more stability for 2S than 2C1S in all testing, except the axial 9 degrees compression tests for shifting. In contrast to the 2C configuration, both modifications (2C1S and 2S) turned out to be stable in dynamic 9 degrees axial compression with a force of 100 up to 1,000 N at 2.5 Hz in 250,000 load cycles. In this in vitro adolescence femur spiral fracture model, the stability of ESIN could be significantly improved by two modifications with additional tension screws. If transferred in clinical practice, these modifications might offer earlier weight bearing and less problems of shortening or axial deviation.