Background Our aim was to evaluate the impact of a computerized echocardiographic simulator on the learning curve for transesophageal echocardiography (TEE) hemodynamic assessment of ventilated patients in the ICU. Methods We performed a prospective study in two university hospital medical ICUs. Using our previously validated skill assessment scoring system (/40 points), we compared learning curves obtained with (interventional group, n = 25 trainees) and without (control group, n = 31 trainees) use of a simulator in the training. Three evaluations were performed after 1 (M1), 3 (M3) and 6 months (M6) while performing two TEE examinations graded by an expert. Competency was defined as a score >35/40. Results Competency was achieved after an average of 32.5 ± 10 supervised studies in the control group compared with only 13.6 ± 8.5 in the interventional group (p < 0.0001). At M6, a significant between-group difference in number of supervised TEE was observed (17 [14–28] in the control group vs. 30.5 [21.5–39.5] in the interventional group, p = 0.001). The score was significantly higher in the interventional group at M1 (32.5 [29.25–35.5] vs. 24.75 [20–30.25]; p = 0.0001), M3 (37 [33.5–38.5] vs. 32 [30.37–34.5]; p = 0.0004), but not at M6 (37.5 [33–39] vs. 36 [33.5–37.5] p = 0.24). Conclusion Inclusion of echocardiographic simulator sessions in a standardized curriculum may improve the learning curve for hemodynamic evaluation of ventilated ICU patients.