Basic concepts of tension stiffening behaviour of reinforced concrete members under tension are reviewed, and different approaches to account for this behaviour are linked together. This includes a "load sharing" approach, where the average load carried by the cracked concrete is used to determine the post-cracking stressstrain response of concrete in tension, and a "tension stiffening strain" approach, which evaluates changes in member stiffness to obtain a reduction in member deformation by including the stiffening effect of the tension carried by concrete between cracks. Shrinkage strains are then included in an analysis of tension stiffening and the results of this analysis are validated with experimental data. The experimental study was carried out for symmetrically reinforced axial tension members having reinforcing percentages between 1% and 2% and shrinkage strain values up to 230 µε. Failure to account for initial member shortening caused by shrinkage leads to an apparent reduction in tension stiffening, which becomes more predominant as the percentage of reinforcement increases. Corrected results indicate that tension stiffening is independent of the reinforcing steel ratio, ρ (within the studied field for 1% < ρ < 2%) and continues to decrease during loading after cracking has stabilized. Test results are also compared with a number of proposals made by other researchers in the past. Prediction of crack widths based on results from axial tension member tests is not significantly affected by shrinkage.Key words: bond, cracking, crack widths, reinforced concrete, shrinkage, stress-strain response, tension, tension stiffening.
