Adding steel fibers to concrete essentially improves its post-crack tensile properties. To determine this experimentally, indirect methods, such as flexural tensile tests, are generally used, which allow only indirect conclusions about the material´s tensile properties. In contrast, direct tensile tests provide the desired result immediately, but are difficult to realize. A key parameter affecting the performance of the SRFC is the orientation of the fibers, which is mainly influenced by the manufacturing process. Typically, when the concrete is cast, the steel fibers align with the edges of the formwork. This is commonly called the wall effect. We address these issues, presenting the setup and results of direct tensile tests on bone shaped specimens with three different steel fiber contents. For each content, a series of specimens with a three-sided formwork (i.e. three-sided wall effect and strong influence on the fiber orientation) and a series with cut-out bones (i.e. one-sided wall effect and less influence on fiber orientation) were fabricated and tested. After these tests, the fiber orientation was determined using an opto-analytical method to quantify the influence of the manufacturing methods on the fiber orientation. Comparing the stress-crack-opening relationships shows that the cut specimens at 0.5 mm crack openings have only about 80% of the tensile strength of three-sided formwork specimens. This effect decreases with larger crack openings and vanishes at about 3 mm crack opening. Finally, a new fiber reinforcement index is defined to correlate observed stress in direct tensile tests to fiber content and orientation in direct tensile tests.