The teaching profession faces a shortage as well as a decline of teaching skills. A possible way to mitigate this is to implement evidence-based induction arrangements. Seventy-one schools with 338 beginning secondary education teachers were randomly allocated to an experimental or a control group. The experimental schools used induction arrangements; the authors measured the effects of these arrangements by using repeated lesson observations and by comparing the rates at which beginners in the control and experimental groups left the teaching profession. Three years later, 14% of the control group and 12% of the experimental group had left. Leaving the profession could be explained by a lack of certification and low initial teaching skill levels. The experimental group exhibited greater improvement in teaching skills compared to the control group. Workload reduction influenced the skill level negatively, and coaching and observing had a strong positive influence on the skill level in Year 3.