Objective: To explore in a multiethnic primary care population the impact of child gender and of race/ethnicity on parent and child reports of school-age anxiety and on the factor structure of the Screen for Childhood Anxiety and Related Emotional Disorders (SCARED). Method: A consecutive sample of 515 children (8 to less than 13 years) and their parent presenting for primary care completed self-report (C) and parent-report (P) versions of the SCARED-41. Results: Neither SCARED scores nor parent-child difference varied significantly with race/ethnicity. Predictors of higher SCARED scores were less parental education, younger child age and female gender. Exploratory factor analysis conducted separately for SCARED-C and SCARED-P yielded four factors. There was large variation in factor structure between SCARED-C and SCARED-P and across ethnic and gender subgroups, greatest for somatic/panic/generalized anxiety and Hispanic children. Conclusions: Primary care triage of anxious children requires data from both the parent and child and must go beyond cross-sectional symptom inventories. Clinicians must elicit from each family their perhaps culturally bound interpretation of the child's somatic and psychological symptoms. (Contains 1 table and 3 figures.)