Components are usually represented by a two state model in conventional generating capacity and composite generation and transmission system reliability studies. Multistate generating unit models create a significant increase in the number of generation contingency states and can result in a considerable increase in the overall solution time. In order to avoid this problem, the derated states are usually amalgamated with the totally forced out state to create the derating-adjusted forced outage rate (DAFOR). This statistic is also known as the equivalent forced outage rate (EFOR). Studies have shown that modeling large generating units in generating capacity adequacy assessments using DAFOR can provide pessimistic appraisals. Many utilities therefore use multistate generating unit representations to assess generating capacity adequacy, in order to obtain more accurate appraisals. There is relatively little published material dealing with the effects of using multistate generating unit representations in composite system adequacy assessment. This paper illustrates these effects by application to the IEEE-reliability test system. Load point and system indices for the test system are presented to illustrate the impact of incorporating multistate representations in composite system adequacy assessment. Attention is focused on the effects of model variations including how many derated states should be used in a multistate model to obtain a reasonably accurate appraisal