Body mass (BM) is a confounding variable in human performance. We hypothesized that adjusting 6-min walk distance (6MWD) for BM differences using allometric scaling would allow meaningful individual and group comparisons. We aimed to use allometric scaling of 6MWD to BM to provide an index for comparing walking performance in middle-aged and older adults. One hundred and twenty subjects (40-87 years) participated. Anthropometry, spirometry, and two walk tests were evaluated. We adjusted 6MWD to BM, gender, and age using an allometric procedure. The allometric exponents were prospectively applied in a validation sample of 44 healthy subjects. Body mass presented significant negative correlation (p < 0.01) with 6MWD · BM(-1) in middle-aged and older adults (r = -0.70 and -0.46, respectively). The allometric exponent (b) for 6MWD was significantly higher for older than middle-aged adults (0.35 ± 0.20 vs. 0.11 ± 0.08, respectively). The resulting BM exponents were similar in male and female subjects (0.22 ± 0.13 and 0.17 ± 0.09, respectively). The correlation between 6MWD · BM(-b ) and BM using the exponents (0.11 or 0.35) was not statistically different from zero (r = 0.00) in the validation sample, suggesting that allometric analysis did not penalize the subjects based on BM. Allometric scaling of 6MWD in middle-aged and older adults may be useful for evaluating walking performance free of the confounding effect of BM, even in the absence of a table of norms.