A previous study found that the modified version of the Pain Assessment Checklist for Seniors with Limited Ability to Communicate (PACSLAC-II) is a valid tool to assess pain in elderly individuals suffering from dementia and who are unable to communicate verbally. The primary objective of this study was to confirm the convergent validity of the PACSLAC-II using direct evaluation of long-term care residents in real-life situations, using two other well-validated pain assessment scales (i.e., PACSLAC and Pain Assessment in Advanced Dementia [PAINAD]). A secondary objective was to document and compare the time required to complete and score each assessment scale. During two potentially painful procedures (i.e., transfer/mobilization), 46 long-term care residents (mean age = 83 ± 10 years) suffering from dementia were observed by three independent evaluators, each using one of the assessment scales (randomly assigned). Correlational analyses and analysis of variance were used to evaluate the association between each scale and to compare scoring time. The PACSLAC (r = 0.61) and the PAINAD (r = 0.65) were both moderately associated with the PACSLAC-II (all p values < .001). The PAINAD's average scoring time (63 ± 19 seconds) was lower than the PACSLAC-II's (96 ± 2 seconds), which was lower than the PACSLAC's (135 ± 53 seconds) (all p values < .001). These results suggest that the PACSLAC-II is a valid tool for assessing pain in individuals with dementia. The time required to complete and score the PACSLAC-II was reasonable, supporting its usefulness in clinical settings.