Newly graduated registered nurses face numerous challenges stemming from high patient workload, complicated interpersonal relationships, and a lack of nursing competence, which can lead to transitional shocks. Clinical judgment and confidence are well‐known keys to successful role transitions for these nurses. Simulation training is proposed as a new modality for enhancing comprehensive clinical competence of nurses, but current evidence on the impact of different simulations on nurses' clinical judgment and confidence are still limited or inconsistent. This study compared the impact of three types of learning modalities on newly graduated registered nurses' clinical judgment, perceptions of self‐confidence, and evaluations of the design features of the learning modalities. A quasi‐experimental design was used. Fifty‐nine participants were randomly assigned to three groups: (1) high‐fidelity simulation, (2) virtual simulation, and (3) case study. Scales were used after the simulation. The virtual simulation group showed a higher level of clinical judgment. The high‐fidelity simulation group felt more confident than the virtual simulation and case study groups. Both the high‐fidelity simulation group and virtual simulation group reported higher scores in the domain of fidelity.