Background: Retronasal olfaction is essential for aroma perception. Its impairment in individuals with olfactory dysfunction (OD) negatively affects dietary behaviour and quality of life. Regular exposure to odours in a form of olfactory training (OT) is used to rehabilitate the sense of smell. However, retronasal recovery following OT remains insufficiently characterised. Objective: To verify whether OT improves retronasal odour identification and whether such improvement is related to orthonasal odour identification recovery. Study Design: Prospective, single-arm clinical study. Methods: Fifty-three patients with OD (predominantly post-COVID-19) completed at least 3 months of OT (conducted bi-daily with four odours). Retronasal function was assessed with the ThreeT test and orthonasal function with the Sniffin’ Sticks battery before and after training. Results: OT significantly improved both retronasal (9.40 ± 0.42 vs. 10.85 ± 0.48; p < 0.001) and orthonasal odour identification (8.75 ± 0.50 vs. 10.53 ± 0.50; p < 0.001). Changes in retronasal performance positively, moderately correlated with changes in orthonasal identification (r = 0.38, p = 0.006). Clinically meaningful retronasal improvement was associated with a higher likelihood of orthonasal recovery, χ²(1) = 5.08, p = 0.024; OR = 4.00 (95% CIs: 1.16–13.83). Conclusions: OT improves retronasal odour identification. The mechanisms of orthonasal and retronasal recovery may be partially overlapping.