ABSTRACT Objective: to compare the acquisition and retention of knowledge about surgical cricothyroidostomy by the rapid four-step technique (RFST), when taught by expository lecture, low fidelity and high-fidelity simulation models. Methods: ninety medical students at UFPR in the first years of training were randomized assigned into 3 groups, submitted to different teaching methods: 1) expository lectures, 2) low-fidelity simulator model, developed by the research team or 3) high-fidelity simulator model (commercial). The procedure chosen was surgical cricothyroidostomy using the RFST. Soon after lectures, the groups were submitted to a multiple-choice test with 20 questions (P1). Four months later, they underwent another test (P2) with similar content. Analysis of Variance was used to compare the grades of each group in P1 with their grades in P2, and the grades of the 3 groups 2 by 2 in P1 and P2. A multiple comparisons test (post-hoc) was used to check differences within each factor (test and group). Statistical significance was considered when p<0.05. Statistical analysis was performed in the statistical software R version 3.6.1. Results: each group was composed of 30 medical students, without demographic differences between them. The mean scores of the groups of the expositive lecture, of the simulator of low fidelity model and of high-fidelity simulator model in P1 were, respectively, 75.00, 76.09, and 68.79, (p<0.05). In P2 the grades were 69.84, 75.32, 69.46, respectively, (p>0.05). Conclusions: the simulation of low fidelity model was more effective in learning and knowledge retention, being feasible for RFST cricothyroidostomy training in inexperienced students.