The Sarbanes–Oxley (SOX) Act was passed in 2002 in response to various instances of corporate malfeasance. The Act, designed to protect investors, led to wide-ranging regulation over various actions of managers, auditors and investment analysts. Part of SOX, and the focus of this study, targeted the disclosure by firms of “pro formaâ€\x9D earnings, an alternate (from GAAP earnings), flexible and unaudited measure of firm performance. Specifically, SOX directed the Securities and Exchange Commission (SEC) to craft regulation which would reduce – and preferably eliminate – any pro forma earnings disclosure which might be “misleadingâ€\x9D. Examining earnings press releases over a 3-year period, this study addresses three questions. Were firms disclosing pro forma in a potentially misleading manner, what was the nature of this potentially misleading disclosure, and did SOX affect the disclosure practices? We find the following. In 2001 (prior to SOX), 53 firms – over 10% of all U.S. S&P 500 firms – were disclosing pro forma earnings in a potentially misleading manner. This was being done most commonly by using traditional GAAP terminology (e.g., “net incomeâ€\x9D) in the press release headline to describe what was later in the press release revealed to be a pro forma amount (i.e., “net income excluding special itemsâ€\x9D). By 2003 (subsequent to the SEC regulation), potentially misleading disclosure practices were seen in less than 1% of the earnings press releases of S&P 500 firms. This significant reduction suggests that managers, prior to the regulation, were either careless in their pro forma reporting practice, or were intentionally – and unethically – attempting to mislead investors. Either way, we conclude that the SEC regulation was both necessary and effective. Copyright Springer Science+Business Media, Inc. 2006pro forma earnings, disclosure management, misleading, regulation,