RATIOS ARE THE BASIC TOOL OF FINANCIAL ANALYSIS. FINANCIAL ANALYSTS FACE TWO MAJOR PROBLEMS WHICH RATIO TO USE OUT OF THEIR NUMEROUS POPULATION AND HOW MUCH RELIABLE ARE THE SELECTED ONES. A REVIEW AND CLASSIFICATION OF THE VARIOUS APPROACHES TO THE FIRST PROBLEM IS ATTEMPTED AND AN EMPIRICAL ANALYSIS OF A SAMPLE OF GREEK RATIOS IS MADE WITH FACTOR ANALYSIS. THE RESULTS ARE CONGRUENT TO THOSEOF PREVIOUS RESEARCHES AND ENLIGHT THE PARTICULARITIES OF THE GREEK DATA. IN THE SECOND PART OF THE STUDY A LARGE NUMBER OF FACTORS AFFECTING THE RELIABILITYOF RATIOS IS ANALYSED AS THE ACCOUNTING DATA STANDARDIZATION, THE ACCOUNTING PROCEDURES AND TECHNIQUES OF MEASUREMENT, VALUATION AND PRESENTATION, INFATIO IMPACT, AND THE MATHEMATICAL NATURE OF RATIOS. EMPHASIS IS PUT ON THEIR STATISTICAL FEATURES AND PARTICULARILY ON THEIR DISTRIBUTIONS. TEN GREEK RATIOS FROM THETEXTILE INDUSTRY ARE ANALYSED FOR A PERIOD OF TEN YEARS. THE MAIN FINDING IS THAT THERE IS NO ONE UNIVERSAL AND EFFECTIVE TRANSFORMATION OF ACCOUNTING DATA TO ACHIEVE NORMALITY. ON THE CONTRARY, THE IDEAL TRANSFORMATION MUST BE FOUND FOR EACH RATIO SEPARATELY (BOX-COX METHOD OF Λ(LAMDA) TRANSFORMATION). NORMALLY AT INDUSTRY LEVE FOUND BY PREVIOUS STUDIES HAS BEEN VERIFIED ONLY FOR THE FINANCIAL STRUCTURE RATIOS.