Accounting practices are flawed. As a consequence, the accounting data generated by firms are generally open to interpretation, often misleading and sometimes patently false. Yet, financial analysts place tremendous confidence in accounting data when appraising investments and investment strategies. The implications of financial analysis based on questionable information are numerous, and range from inexact analysis to acute investment error. To rectify this situation, this paper identifies a set of simple, yet highly effective corrective measures, which have the capacity to move accounting practice into a realm wherein accounting starts to 'count what counts'. The net result would be delivery of accounting data that more accurately reflect firms' economic realities and, as such, are more useful in the task of financial analysis.