Objective: To compare the levels of physical activity (PA) assessed with questionnaires (Leisure Time Physical Activity Instrument [LTPAI], Physical Activity at Home and Work Instrument [PAHWI]) and accelerometry in patients with fibromyalgia; and to analyze the test-retest reliability of these questionnaires. Design: Cross-sectional study. Setting: Local fibromyalgia association. Participants: Participants (N=99; 5 men) with fibromyalgia with a mean age of 50.2 +/- 9.5 years. Interventions: Not applicable. Main Outcome Measures: Participants carried an accelerometer for 1 week and completed the LTPAI and PAHWI twice (separated by a 1-wk interval). The LTPAI and PAHWI were summed to obtain overall values of PA. Results: Time spent in total, moderate, and moderate-vigorous PA was higher (P<.01) when assessed by the LTPAI and PAHWI compared with accelerometry. The Bland-Altman method showed an absence of agreement between the LTPAI and PAHWI and the accelerometer for moderate, moderate-vigorous, and total PA. The test-retest reliability for the workplace subscale and total score of the PAHWI showed high and moderate intraclass correlation coefficients (ICCs), respectively, but also manifested high SE of measurements (up to 179min/d). The LTPAI showed low to moderate ICCs and high SE of measurements (up to 79min/d). For the LTPAI and PAHWI, the ICCs for total activity across the population were low to moderate, and the Bland-Altman method confirmed this lack of agreement. Conclusions: The LTPAI and PAHWI and the accelerometer differ greatly when assessing PA. Furthermore, the LTPAI and PAHWI did not show good levels of test-retest reliability. Therefore, the self-administered LTPAI and PAHWI show questionable usefulness to assess PA in populations with fibromyalgia. (C) 2014 by the American Congress of Rehabilitation Medicine