Tension stiffening and cracking of axial tension members is evaluated for concrete reinforced with steel (reinforcing ratio ρ = 2.0%) and glass fiber reinforced polymer (GFRP) bars (1.3%, 2.0%, and 2.9%), with shrinkage included in the analysis of the member response. Results show that because of a lower bar stiffness the GFRP-reinforced concrete exhibits greater tension stiffening than steel-reinforced concrete for any given value of axial member strain. Transverse cracking in the GFRP-reinforced concrete does not stabilize until much higher values of axial strain are reached, and longitudinal splitting cracks are also evident before cracking has stabilized. Crack widths in concrete reinforced with GFRP bars are larger because of their lower bar stiffness in combination with an increased crack spacing during the crack development stage. Tension stiffening of cracked reinforced concrete is taken into account using an average stress-strain response with a descending branch to model the concrete in tension. A tension stiffening factor is used to characterize this tensile property with an empirical relationship related to the reinforcing bar stiffness and independent of both concrete strength and reinforcing ratio. Results are also compared with the predicted member response based on the 1978 Comité Euro-International du Béton (CEB) CEB-FIP model code approach and American Concrete Institute (ACI) method of using an effective cracked section property for the transformed concrete area. This comparison shows that both methods are valid only for a limited range of reinforcing ratios.Key words: cracking, crack spacing, crack width, GFRP, reinforced concrete, tension stiffening.