Different methods employed for the derivation of the luminosity function in the solar neighbourhood are reviewed. It is shown that most of these methods introduce systematic errors in the results. It is shown that, with certain improvements, van Rhijn’s method is the best one.