Background: Hand hygiene (HH) is a cornerstone of programmes to prevent healthcare associated infections (HAI) globally, but HH interventions are seldom reported from African neonatal units. Methods: We conducted a quasi-experimental study evaluating the impact of a multi-modal intervention (SafeHANDS) on HH compliance rates, alcohol-based handrub (ABHR) usage, the Hand Hygiene Self-Assessment Framework (HHSAF) score, and healthcare-associated bloodstream infection (HA-BSI) rates at a 132-bed South African neonatal unit (4 wards and 1 neonatal intensive care unit [NICU]). The intervention included a campaign logo, HH training, maternal education leaflets, ABHR bottles for staff, and the setting of HH performance targets with feedback. Three 5-month study phases were completed in July 2020 (baseline), December 2020 (early) and May 2021 (intensive). Results: A total of 2430 HH opportunities were observed: 1002 (41.3%) at baseline, 630 (25.9%) at early and 798 (32.8%) at intensive study phases. At baseline, the overall neonatal unit HH compliance rate was 61.6%, ABHR use was 70 mL/patient day, and the baseline HHSAF score was ‘basic’ (165). The overall neonatal unit HH compliance rate was unchanged from baseline to intensive phases (617/1002 [61.6%] vs. 497/798 [62.3%]; p = 0.797). The ABHR use remained similar between phases (70 versus 73 mL/patient day). The HHSAF score improved to ‘intermediate’ level (262). There was no change in the neonatal unit HA-BSI rate. Conclusion: Despite improvement in the HHSAF score, no improvement in overall HH compliance rates, ABHR usage, or HA-BSI rates was observed. Future HH interventions in resource-limited neonatal units should incorporate implementation science and behaviour modification strategies to better understand the barriers and facilitators of HH best practice.