Electronic travel aids (ETAs) can increase the safety and comfort of pedestrians who have a visual impairment by displaying obstacles through a vibrotactile navigation belt. Building a complete picture of relevant obsta- cles and finding a safe route requires ETA users to integrate vibrotactile cues over time and space in their tactile working memory. Previous research suggests that the sense of touch exhibits a working memory that has characteristics similar to vision and audition. However, the capacity of the tactile working memory and the effects of secondary tasks are still under-researched. We investigated tactile working memory capacity of 14 adolescent participants who are blind in an immediate, whole report recall test. Participants received trials consisting of one to five vibration patterns presented sequentially at different locations on their torso rep- resenting obstacles with a direction (vibration location) and distance (vibration pattern). Recall performance was assessed under four conditions: baseline and with distracting background sounds and/or while walking with the long cane. Both walking and ignoring distracting sounds are relevant for everyday use of an ETA and were expected to decrease memory performance. We calculated the 75% correct scores for two memory performance measures: the number of items in a trial (numerosity), and item location and pattern correct. In the baseline condition, the scores were close to ceiling (i.e., 5 items). However, in the presence of distracting sounds and while walking, the scores were reduced to 3.2 items for numerosity and 1.6 items for location and identity correct. We recommend using 2 items as the maximum tactile working memory load in an applied setting unless users are trained and/or can adopt their strategy without unacceptable costs, such as reducing their walking speed.