Virtual Reality (VR) bicycle simulations aim to recreate the feeling of riding a bicycle and are commonly used in many application areas. However, current solutions still create mismatches between the visuals and physical movement, which causes VR sickness and diminishes the cycling experience. To reduce VR sickness in bicycle simulators, we conducted two controlled lab experiments addressing two main causes of VR sickness: (1) steering methods and (2) cycling trajectory. In the first experiment (N = 18) we compared handlebar, HMD, and upper-body steering methods. In the second experiment (N = 24) we explored three types of movement in VR (1D, 2D, and 3D trajectories) and three countermeasures (airflow, vibration, and dynamic Field-of-View) to reduce VR sickness. We found that handlebar steering leads to the lowest VR sickness without decreasing cycling performance and airflow suggests to be the most promising method to reduce VR sickness for all three types of trajectories.