The high-low method is generally less accurate than the least-squares regression method
for analyzing the behaviour of mixed costs.

The high-low method is relatively simple and fast to use, but it relies on only two data points (the highest and lowest), which may not be representative of the entire data set. On the other hand, the least-squares regression method considers all data points and calculates a more accurate estimate of the fixed and variable components of a mixed cost.