Abstract. ABSTRACT Accelerometers measure intensity in an indirect manner, a fact rendering the choice of cut-points (CP) fundamental. The CP for moderate and vigorous intensity of ≥ 1,952 counts is commonly used in studies on older adults, although the criteria for defining the intensities have been established for young subjects. On the other hand, the CP ≥ 1,041 counts, which was developed for older adults, is limited by the lack of maximum stress during the test. The aim of this study was to compare the weekly physical activity of older women using CP ≥ 1,952 counts and ≥ 1,041 counts. The sample consisted of 108 older women who participated in a dance group (n = 69) or in a bingo group (n = 39). The prevalence of active older women varied according to CP, increasing from 56.9% for CP ≥ 1,952 counts to 94.5% for CP ≥ 1,041 counts in the dance group (p<0.05) and from 35.9% to 74.4%, respectively, in the bingo group (p<0.05). Differences were found between groups for weekly step count (dance: 65,386 vs bingo: 46,527, p=0.05; ES: 0.91) and for total physical activity expressed as total count (dance: 3,199,913.2 vs bingo: 2,425,109.7; p=0.03; ES: 0.91). Standardization of accelerometer CP is necessary for future research involving older adults because of the differences found here for the same group.