© 2019 Author(s). Objective: This study aimed to determine the effectiveness of a proficiency-based progression (PBP) training approach to clinical communication in the context of a clinically deteriorating patient. Design: This is a randomised controlled trial with three parallel arms. Setting: This study was conducted in a university in Ireland. Participants: This study included 45 third year nursing and 45 final year medical undergraduates scheduled to undertake interdisciplinary National Early Warning Score (NEWS) training over a 3-day period in September 2016. Interventions: Participants were prospectively randomised to one of three groups before undertaking a performance assessment of the ISBAR (Identification, Situation, Background, Assessment, Recommendation) communication tool relevant to a deteriorating patient in a high-fidelity simulation facility. The groups were as follows: (i) E, the Irish Health Service national NEWS e-learning programme only; (ii) E+S, the national e-learning programme plus standard simulation; and (iii) E+PBP, the national e-learning programme plus PBP simulation. Main outcome measures: The primary outcome was the proportion in each group reaching a predefined proficiency benchmark comprising a series of predefined steps, errors and critical errors during the performance of a standardised, high-fidelity simulation assessment case which was recorded and scored by two independent blinded assessors. Results: 6.9% (2/29) of the E group and 13% (3/23) of the E+S group demonstrated proficiency in comparison to 60% (15/25) of the E+PBP group. The difference between the E and the E+S groups was not statistically significant (¿2 =0.55, 99% CI 0.63 to 0.66, p=0.63) but was significant for the difference between the E and the E+PBP groups (¿2 =22.25, CI 0.00 to 0.00, p<0.000) and between the E+S and the E+PBP groups (¿2 =11.04, CI 0.00 to 0.00, p=0.001). Conclusions: PBP is a more effective way to teach clinical communication in the context of the deteriorating patient than e-learning either alone or in combination with standard simulation.