Peer Reviewed

Lawler's discrepancy model of pay satisfaction was tested to determine its viability as a predictor of pay satisfaction. An analysis of die data seems to indicate that a substantial portion of the model is not supported. Thirteen out of nineteen proposed variables tested were found to have little or no relationships with pay satisfaction. However, social comparison, wage history, status, actual pay, performance and job difficulty were found to have weak to moderately strong relationships with pay satisfaction. Implications for the design of reward systems in organizations are discussed.