BACKGROUND: Nursing home (NH) residents who require assistance during mealtimes are at risk for malnutrition. Supportive handfeeding is recommended, yet there is limited evidence supporting use of a specific handfeeding technique to increase meal intake. OBJECTIVES: To compare efficacy of three handfeeding techniques for assisting NH residents with dementia with meals: Direct Hand (DH), Over Hand (OH), and Under Hand (UH). DESIGN: A prospective pilot study using a within-subjects experimental Latin square design with randomization to one of three handfeeding technique sequences. SETTING AND PARTICIPANTS: 30 residents living with advanced dementia in 11 U.S. NHs. MEASUREMENTS: Time required for assistance; meal intake (% eaten); and feeding behaviors, measured by the Edinburgh Feeding Evaluation in Dementia (EdFED) scale. INTERVENTION: Research Assistants provided feeding assistance for 18 video-recorded meals per resident (N = 540 meals). Residents were assisted with one designated technique for 6 consecutive meals, changing technique every 2 days. RESULTS: Mean time spent providing meal assistance did not differ significantly between techniques. Mean meal intake was greater for DH (67 ± 15.2%) and UH (65 ± 15.0%) with both significantly greater than OH (60 ± 15.1%). Feeding behaviors were more frequent with OH (8.3 ± 1.8%), relative to DH (8.0 ± 1.8) and UH (7.7 ± 1.8). CONCLUSION: All three techniques are time neutral. UH and DH are viable options to increase meal intake among NH residents with advanced dementia and reduce feeding behaviors relative to OH feeding.