Situational Judgment Tests (SJTs) consist of job-related situations and ask participants how they would or should respond to these situations. While SJTs have traditionally been conceptualized as low-fidelity simulations, a recent study fundamentally changed the way we view SJTs. Krumm, Lievens, Hüffmeier, Lipnevich, Bendels, and Hertel (2015) administered SJTs from several construct domains either with or without situation descriptions and revealed that this drastic manipulation did only moderately affect responses to SJT items. Thus, one may conclude that SJTs are not as context-dependent as previously assumed. However, various key questions still remained unanswered. In fact, we do not know how important situation descriptions are for the predictive potential of SJT scores and their relationships with other constructs. To examine this, we administered a SJT on teamwork to 590 participants in a between-subjects design. That is, participants either completed the SJT with or without situations. In addition to the SJT, we assessed participants’ general mental ability, personality, as well as team performance criteria (self- and peer-report). Data were analyzed by inspecting bivariate correlations and along the lines of a multi-group (with vs. without situations) path model with general mental ability and personality predicting SJT performance, which in turn predicted team performance. When constraining path coefficients to identity across both groups, we revealed that only a constraint on the path from conscientiousness to the SJT resulted in a significantly different χ²-square value. All other constraints did not result in significantly different χ²-square values. Comparing bivariate correlations across both groups yielded a similar pattern of results. Thus, we conclude that, by and large, taking away situation descriptions did not substantially affect the SJT’s validity.