The purported mechanism of dowel action in the shear capacity of longitudinally reinforced concrete beams has been revisited, and the results of an experimental investigation of the effect of bar size and amount of tension reinforcement on the load-carrying capacity of such beams designed to fail in shear are discussed. Both the test results, as well as an analysis using the compressive-force path concept, indicate that, for a given grade of steel and concrete, cover and the amount of tension reinforcement, the bar diameter has an insignificant effect on load-carrying capacity, which implies that dowel action is unlikely to be a significant mechanism in the shear capacity of longitudinally reinforced concrete beams. The observed nominal increase in shear strength of the beams with higher amounts of tension reinforcement is attributed to the better control of flexural cracking associated with larger steel ratios, which can increase the component of aggregate interlock in the shear transferred across inclined cracks.