The validity of the two models based on the extended Technology Acceptance Model (Venkatesh and Davies, 2000) in predicting user satisfaction of an incident reporting system in a mandated setting was tested using 84 employees from a large manufacturing company. The models differed in how the construct, subjective norm was represented. Although the results indicated that both models fitted the data, the anticipated relationship between subjective norms and user satisfaction was not supported. Furthermore, some of the antecedent factors did not predict perceived usefulness as anticipated.