# Does the Chow test require the independent variables to be uncorrelated?

The Chow test evaluates if the coefficients in two linear regressions on different data sets are equal.

If two independent variables of the data sets are highly correlated, is the chow test still applicable/reliable?

If not, how much tolerance is acceptable?

In that case, the answer is no. Chow's original paper makes no assumption of any independence between the regressors in the model. The reason for this is that the test is derived conditionally on the observed matrix of regressors, $X$. This isn't effected by partitioning the matrix $X$ at any point.