Few empirical investigations have evaluated learning disabilities (LD) identification methods based on a pattern of cognitive strengths and weaknesses (PSW). This study investigated the reliability of LD classification decisions of the concordance/discordance method (C/DM) across different psychoeducational assessment batteries. C/DM criteria were applied to assessment data from 177 second-grade students based on 2 psychoeducational assessment batteries. The achievement tests were different, but were highly correlated and measured the same latent construct. Resulting LD identifications were then evaluated for agreement across batteries on LD status and the academic domain of eligibility. The 2 batteries identified a similar number of participants as having LD (80 and 74). However, indices of agreement for classification decisions were low (? = 0.29), especially for percent positive agreement (62%). The 2 batteries demonstrated agreement on the academic domain of eligibility for only 25 participants. Cognitive discrepancy frameworks for LD identification are inherently unstable because of imperfect reliability and validity at the observed level. Methods premised on identifying a PSW profile may never achieve high reliability because of these underlying psychometric factors. An alternative is to directly assess academic skills to identify students in need of intervention.