BACKGROUND Physical activity (PA) assessment by questionnaire is a cornerstone in the field of sport epidemiology studies. The Physical Activity Questionnaire for Children (PAQ‐C) has been used widely to assess PA in healthy school populations. The aim of this study was to evaluate the reliability and validity of the PAQ‐C questionnaire in Spanish children using triaxial accelerometry as criterion. METHODS Eighty‐three (N = 46 boys, N = 37 girls) healthy children (age 10.98 ± 1.17 years, body mass index 19.48 ± 3.51 kg/m2) were volunteers and completed the PAQ‐C twice and wore an accelerometer for 8 consecutive days. Reliability was analyzed by the intraclass correlation coefficient (ICC) and the internal consistency by the Cronbach's α coefficient. The PAQ‐C was compared against total PA and moderate to vigorous PA (MVPA) obtained by accelerometry. RESULTS Test‐retest reliability showed an ICC = 0.96 for the final score of PAQ‐C. Small differences between first and second questionnaire administration were detected. Few and low correlations (rho = 0.228‐0.278, all ps < .05) were observed between PAQ‐C and accelerometry. The highest correlation was observed for item 9 (rho = 0.311, p < .01). CONCLUSIONS PAQ‐C had a high reliability but a questionable validity for assessing total PA and MVPA in Spanish children. Therefore, PA measurement in children should not be limited only to self‐report measurements.