The SEC comment letter is the correspondence between SEC staff and SEC filers about the filers' public information disclosures. The intensity of comment letters in terms of the use of strong/weak modal language can reflect perceived deficiencies in the reviewed filings. This paper uses text mining to examine the intensity of SEC comment letters. A measure of intensity based on the modality of comment letters is developed. Empirical analysis is conducted on a sample of initial comment letters related to 10-K filings. Results show that intensity is positively associated with the probability of a restatement of the reviewed 10-K filings and that this association is robust using both the original Loughran and McDonald (2011) word lists and the modified word lists.