Current guidelines require hand hygiene before donning nonsterile gloves, but evidence to support this requirement is lacking. To evaluate the effectiveness of a direct-gloving policy on adherence to infection prevention practices in a hospital setting. This mixed-method, multicenter, cluster randomized clinical trial was conducted at 4 academic centers in Baltimore, Maryland, or Iowa City, Iowa, from January 1, 2016, to November 30, 2017. Data analysis was completed April 25, 2019. Participants were 3790 health care personnel (HCP) across 13 hospital units. Hospital units were randomly assigned to direct gloving, with hand hygiene not required before donning gloves (intervention), or to usual care (hand hygiene before donning nonsterile gloves). The primary outcome was adherence to the expected practice at room entry and exit. A random sample of HCPs' gloved hands were imprinted on agar plates at entry to contact precautions rooms. The intention-to-treat approach was followed, and all analyses were conducted at the level of the participating unit. Primary and secondary outcomes between treatment groups were assessed using generalized estimating equations with an unstructured working correlation matrix to adjust for clustering; multivariate analysis using generalized estimating equations was conducted to adjust for covariates, including baseline adherence. In total, 13 hospital units participated in the trial, and 3790 HCP were observed. Adherence to expected practice was greater in the 6 units with the direct-gloving intervention than in the 7 usual care units (1297 of 1491 [87%] vs 954 of 2299 [41%]; P < .001) even when controlling for baseline hand hygiene rates, unit type, and universal gloving policies (risk ratio [RR], 1.76; 95% CI, 1.58-1.97). Glove use on entry to contact precautions rooms was also higher in the direct-gloving units (1297 of 1491 [87%] vs 1530 of 2299 [67%]; P = .008. The intervention had no effect on hand hygiene adherence measured at entry to non-contact precautions rooms (951 of 1315 [72%] for usual care vs 1111 of 1688 [66%] for direct gloving; RR, 1.00 [95% CI, 0.91-1.10]) or at room exit (1587 of 1897 [84%] for usual care vs 1525 of 1785 [85%] for direct gloving; RR, 0.98 [95% CI, 0.91-1.07]). The intervention was associated with increased total bacteria colony counts (adjusted incidence RR, 7.13; 95% CI, 3.95-12.85) and greater detection of pathogenic bacteria (adjusted incidence RR, 10.18; 95% CI, 2.13-44.94) on gloves in the emergency department and reduced colony counts in pediatrics units (adjusted incidence RR, 0.34; 95% CI, 0.19-0.63), with no change in either total colony count (RR, 0.87 [95% CI, 0.60 to 1.25] for adult intensive care unit; RR, 0.59 [95% CI, 0.31-1.10] for hemodialysis unit) or presence of pathogenic bacteria (RR, 0.93 [95% CI, 0.40-2.14] for adult intensive care unit; RR, 0.55 [95% CI, 0.15-2.04] for hemodialysis unit) in the other units. Current guidelines require hand hygiene before donning nonsterile gloves, but evidence to support this requirement is lacking. The findings from this cluster randomized clinical trial indicate that a direct-gloving strategy without prior hand hygiene should be considered by health care facilities. ClinicalTrials.gov Identifier: NCT03119389.