AbstractThe widely used shear-lag theory for the stresses and displacements in a short fibre composite is discussed and compared with more exact theories in the literature. A simple modification of shear-lag theory, using stress functions, is proposed that allows prediction of radial stresses in the fibre, while being easy to use for calculations. Predictions from the present theory are compared with those from the shear-lag theory and with some of the more exact theories. At low fibre concentrations, the present theory underestimates the radial stress at the fibre ends while predicting shear stresses comparable with those from shear lag. At high fibre concentrations both radial and shear stresses are overestimated.MST/635