Abstract Background Neuropsychiatric symptoms (NPS) are features of Alzheimer’s Disease (AD) and a main cause of institutionalization. NPS in AD include agitation, depression, anxiety, elation, apathy, disinhibition, irritability, delusions, hallucinations and aberrant motor behavior. The onset of NPS worsens functional deficits by accelerating patients' cognitive decline and in 80% of cases increases caregiver burden and distress. NPS are difficult to treat and currently used medications have black‐box warnings, and weak evidence of efficacy in dementia patients. Here, we present preliminary data for improvement in NPI scores and NPI Caregiver Distress (NPI‐D) scores using IGC‐AD1, a combination of tetrahydrocannabinol (THC) and melatonin (IGC‐AD1). Method Twelve patients with mild (15.38%) to moderate (84.6%) AD (NIA‐criteria, 10‐active, 2‐placebo, 81.5±5.5yrs, 69.2% women) participated in a three Cohort Phase‐1, MAD, safety and tolerability trial (IND146069, NCT04749563). In Cohort‐1, IGC‐AD1 was administered QD at 1ml for 14‐days (EOT). In Cohorts‐2 and 3, one ml BID and TID were administered respectively with a minimum of 4‐days washout between Cohorts. For all three Cohorts the NPI / NPI‐D were administered at baseline and EOT of each Cohort. Daily, solicited and non‐solicited adverse events (AEs) were monitored along with vital signs. A Wilcoxon matched‐pairs signed‐rank test (R‐Studio, dplyr ) was used to independently compare differences between mean NPI and NPI‐D scores at baseline and EOT. Result In Cohort‐1 mean NPI at baseline and EOT was 31.5 and 14.8 respectively (mean difference = ‐16.7, v = 55, p = 0.003). In Cohort‐2 mean NPI at baseline and EOT was 22.2 and 12.4 respectively (mean difference = ‐9.8, v = 48, p = 0.021). In Cohort‐3 mean NPI at baseline and EOT was 16 and 7.9 respectively (mean difference = ‐8.1, v = 35, p = 0.010). We report similar improvements in NPI‐D. (Cohort‐1: mean difference = ‐7, v = 15, p = 0.031; Cohort‐2: mean difference = ‐6.3, v = 43, p = 0.009; Cohort‐3: mean difference = ‐3.4, v = 30, p = 0.053). No changes in concomitant medication were observed. No serious AEs, no deaths, and no dropouts due to AEs were reported. Conclusion GC‐AD1 was safe and well tolerated and is potentially efficacious in reducing NPS and associated caregiver distress, clinically and statistically, in AD as measured by the NPI, warranting a larger placebo‐controlled study.