This research aims to develop and test the effectiveness of digital literacy instruments on students' learning outcomes. This study is a mixed method that combines survey and pre-experimental research. The survey research was conducted through a quantitative approach using a questionnaire as the research instrument with assessed numerical items. The survey respondents consisted of 318 students from various universities in Indonesia who were taken through cluster random sampling technique. The implementation of the instrument was carried out through a pre-experimental design to 103 students to test the instrument's effectiveness. The confirmatory factor analysis result shows that the model forms three factors, which are factor 1 (communicating digital content), factor 2 (exploring digital content), and factor 3 (creating and using digital content). The result of the goodness of fit model shows that the instrument has met all the criteria with a value of X[superscript 2]/df = 1.642 (<3.00), RMSEA = .079 (= 0.08), GFI = .900 (= 0.90), AGFI = 0.825 (= 0.90), TLI = 0.939 (= 0.90), and CFI = .951 (= 0.90). Empirically the instrument has good discrimination power so that it can measure students' digital literacy skills at different academic ability levels (Sig. <0.05). In addition, the instrument has been effectively used to measure and predict student learning outcomes (R square = 0.255, Sig. < 0.05). The instruments' result is relevant to current conditions and can further explore students' digital literacy skills.