There is no agreement among authors that study fish condition by the allometric method (K=W/Lb) with regard to the best procedure for b coefficient calculation. Some authors use a constant coefficient for all sub-samples (seasons of the year, for instance), whilst others calculate b value for each sub-sample. To demonstrate which of these methods fits better, this study verified that the use of one b value for each sub-sample leads to distortion of Condition Factor values. Comparing the two tested methods, it may be concluded that the method which calculates b coefficient from groupings of all individuals and uses b as a constant value for all sub-samples is the most convenient method to study fish condition.