Congestion in bicycle traffic is a daily occurrence at many urban intersections. It is known that a higher density in the queue leads to a higher discharge rate. In theory, higher jam densities than those currently observed in practice are feasible. This leads to our hypothesis that the delay at intersections can be further reduced when cyclists are encouraged to queue up closer together. To explore this option, we carried out an experiment in which the queue configuration was influenced to increase the jam density. This paper presents ways to increase the queuing density, up to twice the density found without instructions. Results show that increasing the jam density does indeed increase the queue discharge rate; this also holds for jam density values that exceed those observed in normal queuing conditions. The efficiency of the queue discharge process, captured by the discharge rate, was found to increase by 40% when cyclists queue up closely together. Qualitative comparison of the queuing positions and discharge patterns showed that the discharge sequence is largely determined by the queuing position, and that cyclists keep a distance from each other in both time and space during the queue discharge phase. When applied in practice, these findings can be used to update the signal length and green phases for all traffic, thereby reducing congestion in urban areas.@en