Order picking is one of the most important processes in warehouse oper- ations and deals with the preparation of customer orders. In picker-to-parts systems, pickers navigate around the warehouse collecting the required prod- ucts to prepare the orders. The joint order batching and picker routing problem (JOBPRP) is to determine simultaneously how the orders are grouped, and the route that each picker will follow. In general for the picking process, works in the literature do not consider congestion effect when multiple pickers work simultaneously. In this work, we introduce the JOBPRP including picker con- gestion and propose a heuristic based on a column generation algorithm to solve it. In practical terms, congestion can be conceptualized as a situation produced when two or more pickers are competing for the same space during the same time, having as a consequence a delay in the travel time. To quantify the delay produced by congestion we propose to divide the planning horizon into homogeneous time intervals, and for each of them the number of pickers visiting each sub-aisle is computed. When two or more pickers are visiting the same sub-aisle during the same time interval they incur a delay in their travel time. To avoid congestion, picker routes have to be synchronized. We propose an extended formulation for the JOBPRP including congestion in which variables represent possible picking routes. Each route has information about the time in which each product is picked and the level of congestion of each sub-aisle. The model is solved using a heuristic column generation approach based on the resolution of the linear relaxation of the extended formulation. Several computational experiments were performed on instances from the literature and a new set of instances produced to evaluate scenarios with high congestion