Background and objectives Injection and inhalational heroin use are associated with different levels of brain exposure to heroin and its metabolites and differences in the severity of dependence, which might lead to differential impacts on neuropsychological functions. We examined the difference and the magnitude of difference in the neuropsychological functions between inhalational and injection heroin-dependent subjects and also compared them with healthy controls. Methods The study sample comprised three groups: 73 subjects with injection heroin dependence, 74 with inhalational heroin dependence, and 75 healthy controls (HC). We excluded patients with HIV, head injury, epilepsy, and severe mental illness. Neuropsychological assessments were done by Standard Progressive Matrices, Wisconsin Card Sorting Test (WCST), Iowa Gambling Task, Trail-Making Tests A and B (TMT), and Verbal and Visual Memory 1 and 2 Backtests (NBT). We estimated independent effects of the groups on various neuropsychological test parameters, adjusted for age and duration of dependence. Results In the WCST, the inhalational heroin-dependent group took more trials to complete the first category and had higher scores in the failure to maintain set than controls. The intravenous group had higher total errors than controls in verbal working memory tests and Visual Working Memory 2 Backtest. This group scored higher commission errors in the Verbal 2 Backtest than the controls. The two groups of heroin users differed in failure to maintain set and Verbal Working Memory 2 Backtests. The effect sizes of the group differences were modest. Conclusion and scientific significance Either route of heroin use is associated with cognitive impairments; inhalational and injection use involve different cognitive domains.