BACKGROUND Quadriceps strength and size are commonly reduced in chronic obstructive pulmonary disease (COPD). We wished to assess volitional and nonvolitional ankle dorsiflexor strength in COPD. METHODS Quadriceps and ankle dorsiflexor strength were measured by maximum voluntary contraction (MVC) and by twitch responses to supramaximal femoral and fibular nerve stimulation. Cross-sectional areas of the tibialis anterior (TA(CSA)) and rectus femoris muscles (RF(CSA)) were measured by ultrasound. RESULTS Eighteen elderly subjects and 20 COPD patients [mean(SD) %predictedFEV(1) 50(20)%] participated. No significant difference in fat-free mass index, ankle dorsiflexor strength, or TA(CSA) were observed in the presence of reduced quadriceps strength and size in COPD [mean MVC difference: -10.9 kg (95% confidence interval {CI}: -17.1 kg to -4.8 kg, P < 0.01; mean RF(CSA) difference -119 mm(2), 95% CI: -180 mm(2) to -58 mm(2), P < 0.01)]. CONCLUSIONS Ankle dorsiflexor strength is less attenuated than quadriceps strength in COPD patients with moderate airflow obstruction. Direct quadriceps assessment may be more relevant than measurement of lower limb fat-free mass.