Although the "digital natives" theory is still used to justify the implementation of technologies such as video games in the language classroom, whether digital nativeness (i.e. the extent to which an individual embodies the characteristics attributed to digital natives) may be used as a predictor of digital game-based learning outcomes is to be analysed. An empirical study was conducted to test the predictive strength of some attributes of digital nativeness, as determined by Teo's (2013) "Digital Native Assessment Scale," on L2 vocabulary acquisition and on educational engagement. A vocabulary learning game was played by 86 university EFL students, who additionally completed post- and delayed knowledge tests on the vocabulary contained in the game, the DNAS questionnaire, and questions on educational disengagement and perceived need for technology-related changes. The regression analysis revealed that the factor "able to multitask" was a predictor of poorer L2 vocabulary acquisition, whereas the factors "comfortable with technology" and "thrive on instant gratifications and rewards" had a positive impact on learning outcomes. Additionally, the analysis showed that none of the attributes of digital natives that make up the DNAS could predict either students' disengagement with the educational system or their perceived need for technology-based changes in education.