TwinLife, an ongoing German study of twins and their families, investigates cognitive performance as one factor among many that contribute to the development of social inequality. Participants completed the CFT 20-R, a nonverbal intelligence assessment. The current analysis applied a two-parameter logistic item response theory model using Mplus software to subtest results from twin pairs in the three oldest birth cohorts, ranging in age from 10 to 25 years old. The findings indicated that the 2PL model fit the data considerably better than the one-parameter logistic model did for all four of the CFT 20-R subtests used in TwinLife. Results from the 2PL model, including item and person parameters and test information, are discussed. In addition, the items were assessed for measurement invariance across age cohort and gender. Fit statistics reveal little difference in item function according to these demographic factors, meaning that the CFT 20-R may be valid in heterogeneous samples.