This preliminary study examines the utility of the Community Engagement Commitment Scale and reports on factors that predict student-teachers' commitment to community engagement in their second year of study at a South African university. Independent variables included Dalbert's personal belief in a just world, an adaptation of Rusbult et al.'s original predictors of "Commitment level" ("Satisfaction," "Quality of alternatives," "Investment size") and a new factor, "Meaningfulness." The latter factor is based theoretically on the work of Lieberman and Heine, Proulx and Vohs and was included to study the extent to which identity-expression could be related to "Commitment level." We administered the Community Engagement Commitment Scale on a non-random sample of 414 student-teachers in their second year of study, who all participated in compulsory community engagement project. The results support a four-factor model of Commitment with "Meaningfulness," "Quality of alternatives" and "Investment as predictors of Commitment level." Surprisingly, all the "Satisfaction" items loaded with the "Commitment level" items on one factor. Together, "Meaningfulness," "Quality of alternatives" and "Investment" accounted for approximately two-thirds of the variance in "Commitment level." Significant (p?less than 0.01) language group differences were also observed on "Meaningfulness" and "Commitment level." The results of the study suggest the importance of studying meaningfulness as an additional predictor of "Commitment level" in community engagement contexts. (Contains 4 tables.)