The three mechanisms governing the strand-concrete (S-C) bond in pretensioned concrete (PTC) systems are: (i) adhesion, (ii) mechanical interlock, and (iii) friction. These mechanisms can be influenced by various parameters like the compressive strength of concrete (fc), applied prestress (fps), and embedment length of the strand (le). The existing test methods for S-C systems use unstressed strands and defines bond strength (τb), as the stress corresponding to a slip of 2.5 mm at free end (say, τ2.5). The unstressed strands and τ2.5 approach may lead to unconservative results and significant scatter. This paper presents the experimental program on the effects of fc (43 and 62 MPa), fps (0.1 fpu and 0.7 fpu), and le (500 and 1000 mm) on the bond stress-slip (τ-s) behaviour in PTC systems. Analysis of the results of tests on 24 specimens indicated that (i) defining the τb as τ2.5 at the free end is not suitable for PTC specimens and (ii) the τ2.5 at live end exhibits significant scatter and is dependent on the embedment length, le. Consequently, the paper proposes a conceptual model on strand-concrete bond behavior and a rational method to determine τb as the stress corresponding to the yield bond stress (τyield). The determined τyield is independent of fps and le and exhibits less scatter than τ2.5. Finally, the use of taut strand specimens is recommended as they are easier to prepare than stressed strand specimens, to determine τyield. © 2021