A matrix is said to have strongly linearly independent columns (or, in the case of square matrices, to be strongly regular) if for some vector b the system A x = b has a unique solution. We formulate a necessary and sufficient condition for a linear system of equations over a fuzzy algebra to have a unique solution and prove the equivalence of strong regularity and trapezoidal property. Moreover, an algorithm for testing these properties is reviewed.