ObjectivesTo assess the construct validity and responsiveness of 3 performance-based tests in individuals with knee osteoarthritis (KOA). DesignThis study has a prospective cohort design. SettingAssessments were administered at the Federal University of São Carlos (Brazil) by 3 different raters. ParticipantsA total of 107 participants with KOA were aged ≥40 years (N=107). InterventionsNot applicable. Main Outcome MeasureThis study assessed the construct validity and responsiveness of 3 performance-based tests: 40-meter fast-paced walk test (40-m FPWT), 11-step stair climb test (11-step SCT), and 30-second chair stand test (30-s CST). The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), Knee Injury and Osteoarthritis Outcome Score (KOOS), sedentary behavior (activPAL), and numeric pain rating scale (NPRS) were used as comparator instruments. Measurements were obtained at baseline and after 6 months. The construct validity and responsiveness were calculated using Spearman correlation coefficient within predefined hypotheses based on a panel comprising 5 experts in the field of outcome measurement in KOA. ResultsThe three performance-based tests showed excellent (ICC>0.9) reliability (intra and inter); however, only the 40-m FPWT and 11-step SCT were considered valid and responsive because both confirmed >75% of the priori hypotheses. The 30-s CST was not considered valid or responsive because it has not confirmed ≥75% of the priori hypotheses. ConclusionThe 3 performance-based tests are reliable for intra and inter assessments. The 40-m FPWT and 11-step SCT are both valid and responsive for measuring physical function in patients with KOA. The 30-s CST was not considered valid or responsive. Therefore, 40-m FPWT and 11-step SCT can detect changes over time and are indicated for functional assessment in clinical practice.