Humans are known to construct cognitive maps of their everyday surroundings\nusing a variety of perceptual inputs. As such, when a human is asked for\ndirections to a particular location, their wayfinding capability in converting\nthis cognitive map into directional instructions is challenged. Owing to\nspatial anxiety, the language used in the spoken instructions can be vague and\noften unclear. To account for this unreliability in navigational guidance, we\npropose a novel Deep Reinforcement Learning (DRL) based trust-driven robot\nnavigation algorithm that learns humans' trustworthiness to perform a language\nguided navigation task. Our approach seeks to answer the question as to whether\na robot can trust a human's navigational guidance or not. To this end, we look\nat training a policy that learns to navigate towards a goal location using only\ntrustworthy human guidance, driven by its own robot trust metric. We look at\nquantifying various affective features from language-based instructions and\nincorporate them into our policy's observation space in the form of a human\ntrust metric. We utilize both these trust metrics into an optimal cognitive\nreasoning scheme that decides when and when not to trust the given guidance.\nOur results show that the learned policy can navigate the environment in an\noptimal, time-efficient manner as opposed to an explorative approach that\nperforms the same task. We showcase the efficacy of our results both in\nsimulation and a real-world environment.\n