Objective/Hypothesis A common, lasting condition from traumatic brain injury is impairment to smell. In patients with olfactory impairment, recent meta‐analyses have demonstrated that olfactory training consistently improves higher‐order functions, such as odor identification. The focus of this work was to assess effects of olfactory training (OT) in posttraumatic olfactory loss patients through several metrics including psychophysical, olfactory bulb (OB) volume, and functional magnetic resonance imaging. Study Design Prospective cohort study. Methods Sniffin’ Sticks were used to classify two patient groups (anosmic N = 23 and hyposmic N = 14) and measure changes after OT. Additionally patients were asked the intensity, valence, and uncued identification of odors presented (coffee and peach) within the magnetic resonance imaging scanner before and after olfactory training. Olfactory training was performed twice daily with a four‐odor training set for 24 weeks, and sets were replaced halfway through the entire training session (~12 weeks). Results Patients had an increase in test scores (threshold and identification) and in‐scanner intensity ratings and identification. Anosmic patients showed improved olfactory thresholds to 2‐phenylethanol, increased intensity ratings, and activation in the right superior frontal gyrus (SFG) to odors after OT. Hyposmic patients were able to identify odors better after training. This behavior was mirrored with increased, ipsilateral activations in semantic processing areas such as Broca's area, left angular gyrus, and left SFG. Conclusions Taken together, along with neither patient group showing changes in OB volumes, OT improves olfactory performance in patients with posttraumatic olfactory loss and seems to be driven, at least in part, by top‐down processes (central) rather than bottom‐up (peripheral). Levels of Evidence 2Laryngoscope, 129:1737–1743, 2019