PURPOSE: To compare parent-reported child physical activity (PA) with accelerometer-measured in a sample of racially/ethnically diverse families and examine whether self-reported and objective measures differed by sex. METHODS: 146 parent-child dyads, with children (5-7 years old; 47% girls) represented by six racial/ethnic groups (n=23 African American, n=25 American Indian, n=25 Hispanic/Latino, n=25 Hmong, n=24 Somali, and n=24 White) were included in this analytic sample. Parents reported average weekly hours of child light (LPA), moderate (MPA), and vigorous PA (VPA) using a validated self-report measure. Child LPA, MPA, and VPA were objectively measured by accelerometry (ActiGraph GT1M) for 8 days. Correlation analyses were used to compare self-reported and accelerometer-measured PA variables. Multiple linear regression assessed the association between parent self-reported child LPA, MPA and VPA and accelerometry, and whether child sex moderated these associations. All models were adjusted for child age, sex, BMI, race/ethnicity, and household income. RESULTS: The average weekly hours of LPA, MPA, and VPA measured by parent self-report and accelerometry were 3, 2.8, 2.3, and 30±6.6, 3.9±1.5, 1.6±1.1 hours per week, respectively. Pearson correlations between self-reported and accelerometer-measured LPA, MPA, and VPA were 0.03, 0.25, and 0.22, respectively. The relationship between self-reported and objective PA were similar between girls and boys (all P-interaction > 0.1). Accelerometer-measured MVPA revealed 23.8% of children met the U.S. PA guidelines. CONCLUSION: Overall, self-reported and objective measures of PA were poorly correlated in this racially/ethnically diverse sample of parent-child dyads. Misunderstanding of what constitutes different levels of child PA by parents, the survey tool, and the sporadic nature of PA behaviors in young children might explain these findings. When accelerometry is not available, future studies should incorporate descriptive measures of different types of PA and choose a different survey tool, such as Ecological Momentary Assessment to have better estimates of parent-reported child PA throughout the day. Interventions for PA promotion among racially/ethnically diverse children are needed given the low levels of MVPA in this sample.