In a recent publication [Biochem. Biophys. Res. Commun. 317 (2004) 957] it was proposed that the 'effective number of codons' ((N) over capc) in a gene should be calculated by summing the individual amino acid (N) over capc's using rounding whenever the codon homozygosities are lower than the reciprocal value of the number of members of the synonymous families. This led Marashi and Najafabadi to examine the consequences of individual re-adjustment when comparing observed (N) over capc with the expected Nc under assumptions of no selection, and C = G and A = T [Biochem. Biophys. Res. Commun. 324 (2004) 1]. Clearly, the present methodology has some weaknesses; in this work, I discuss these in relation to the observations by Marashi and Najafabadi, and finally an alternative method for the calculation of (N) over capc is introduced with the purpose of eliminating the need for re-adjustments. (C) 2004 Elsevier Inc. All rights reserved.