The authors examined the validity of a method commonly used to test alternative hypotheses regarding the causes of comorbidity: the examination of underlying deficits of comorbid disorders. The authors simulated data in which the true causes of comorbidity were known, then compared the patterns of underlying deficits of the comorbid disorders found in the simulated data with the predicted results. The method of examining the underlying deficits of comorbid disorders could distinguish between several comorbidity models, including those that could not be distinguished well using other methods. The ability to distinguish the correct model decreased as the sample size and the correlation between the underlying deficits and the symptom scores decreased, suggesting that the issue of power should be considered carefully.Keywords: comorbidity, comorbidity models, comorbid disorders, underlying deficits, simulations Epidemiological evidence suggests that the rate of comorbidity among psychiatric disorders exceeds the rate expected by chance (e.g., Newman et al., 1996;Simonoff et al., 1997). Many researchers have considered alternative hypotheses for the causes of comorbidity and ways to test these hypotheses