OBJECTIVE: To evaluate the retest reliability and quantify the degree of measurement error when measuring isometric muscle strength with a hand-held dynamometer for people with chronic obstructive pulmonary disease (COPD). DESIGN: Retest reliability of hand-held dynamometry for 4 muscle groups was assessed on 2 occasions separated by a 2-week interval. SETTING: Community rehabilitation center. PARTICIPANTS: Eight men and 4 women (mean age +/- standard deviation, 71.4+/-10.3y) with moderately severe COPD (percentage of predicted forced expiratory volume in 1 second, 41.5%+/-17.7%). INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Muscle strength (in kilograms). Statistical analysis was conducted by calculating intraclass correlation coefficients and 95% confidence intervals for both group and individual scores. RESULTS: All reliability coefficients were greater than .79. Muscle strength would need to increase by between 4% and 18% in groups of people with COPD and between 34% and 58% in a person with COPD to be 95% confident of detecting real changes. CONCLUSIONS: Hand-held dynamometry is suitable for monitoring change in muscle strength and testing hypotheses for groups of people with COPD. However, hand-held dynamometry is not likely to detect changes in muscle strength for a person with COPD.