Contribution: This paper shows that identification with engineering for engineering graduate students is positively and significantly predicted by engineering interest, competence, recognition, and interpersonal skills competence. Background: Prior studies of engineering identity on undergraduates identified several factors (e.g., engineering interest and engineering recognition) as positive predictors of identification of engineering. Engineering competence, achieved by participating in design projects, is a crucial part of students' efforts to become more innovative engineers. Identity theory is used to understand undergraduates' persistence in engineering, as students with stronger engineering identification are more likely to persist. More work is needed focusing on graduate students. Research Questions: Do engineering identity measurement frameworks studied for undergraduate students also apply to graduate students? Do they correlate with intention to complete the degree? What predicts the engineering identity of engineering master's and doctoral students? Methodology: Interviews informed development and adaptation of a multi-scale survey instrument. Factor analyses identified four factors that relate to graduate engineering identity: 1) engineering interest; 2) engineering recognition; 3) engineering competence; and 4) interpersonal skills competence. Three sequential multiple linear regression models were used to predict engineering graduate students' engineering identity. Findings: The final regression model, which includes student characteristics and the four factors resulting from confirmatory factor analysis, predicts 60% of the variance in engineering identity--substantially more than similar undergraduate engineering identity models. All four factors were significant and positive predictors of graduate students' engineering identity. The engineering recognition factor in particular needed adaptation to emphasize peers and faculty members over family, although family remained important.