The traditional measure of river-channel sinuosity, P, is shown to var
y with scale of measurement, the geometric character of the particular
reach, and the starting point of the divider walk. Meander wavelength
, L, suffers from non-reproducibility caused by the subjective identif
ication of inflection points, a problem compounded by the irregularity
of river meanders. Applicability of the angle measure technique (AMT)
to the measurement oi meander planform is demonstrated. The AMT consi
ders variations in The complexity of meandering with scale, eliminatin
g the dependence of results on scale of measurement. The AMT also samp
les the river at random points rather than sequentially unlike the div
ider walk used to calculate P,avoiding the dependence of P on the star
ting position of the divider walk. The AMT also does not. require subj
ective identification of inflection points as wavelength L does. Addit
ionally, the traditional measures P and L are difficult to apply to ma
nifestly underfit rivers. In this study, the AMT is applied to two und
erfit rivers and measures of meander size and sinuosity are calculated
for each set of meanders.