Widespread use of the Modern Language Aptitude Test (MLAT) in L2 studies of individual differences implicitly assumes that L2 aptitude is a distinct cognitive facet. There is considerable evidence for prediction from L1 abilities to L2 learning. In this longitudinal study, L1-MLAT-L2 relations were examined in 307 US secondary students based on six L1 and six L2 measures of language and literacy, and the MLAT. Mediation and regression analyses revealed that each L1 measure individually predicted all L2 scores and MLAT; the L1 measures collectively substantially predicted MLAT scores; MLAT is a significant but moderate mediator of prediction from L1 to L2 scores; and prediction from MLAT to L2 scores is significantly and substantially due to variance in L1 abilities captured by MLAT. Overall, prediction from MLAT is due primarily to its functioning as a measure of L1 abilities, although substantial L1 variance which predicts L2 scores is not captured by the MLAT.