The Association of American Medical Colleges has encouraged educators to investigate proper linkage of simulation experiences with medical curricula. The authors aimed to determine if student knowledge and satisfaction differ between participation in web-based and manikin simulations for learning shock physiology and treatment and to determine if a specific training sequencing had a differential effect on learning. All 40 second-year medical students participated in a randomized, counter-balanced study with two interventions: "group 1" (n = 20) participated in a web-based simulation followed by a manikin simulation and "group 2" (n = 20) participated in reverse order. Knowledge and attitudes were documented. Mixed-model ANOVA indicated a significant main effect of time (F[subscript 1,38] = 18.6, P less than 0.001, [[eta][subscript [rho]]][superscript 2] = 0.33). "Group 1" scored significantly higher on "quiz 2" (81.5%) than on "quiz 1" (74.3%, t[subscript 19] = 3.9, P = 0.001), for an observed difference of 7.2% (95% confidence interval: 3.3, 11.0). Mean quiz scores of "group 2" did not differ significantly ("quiz 1": 77.0% and "quiz 2": 79.7%). There was no significant main effect of group or a group by time interaction effect. Students rated the simulations as equally effective in teaching shock physiology (P = 0.88); however, the manikin simulation was regarded as more effective in teaching shock treatment (P less than 0.001). Most students (73.7%) preferred the manikin simulation. The two simulations may be of similar efficacy for educating students on the physiology of shock; however, the data suggest improved learning when web-based simulation precedes manikin use. This finding warrants further study. (Contains 4 figures and 1 table.)