Completeness of tuberculosis case reporting in Puerto Rico was assessed. Cases diagnosed among hospitalized, tuberculosis, and human immunodeficiency virus clinic patients during 1992 were retrospectively reviewed. Hospital discharge diagnoses, pharmacy listings of patients receiving anti-tuberculous medications, laboratory and acquired immunodeficiency syndrome registry data were used for case finding in selected hospitals and clinics. Identified cases were matched to the health department TB case registry to determine previous reporting through routine surveillance. Records of unreported cases were reviewed to verify tuberculosis diagnoses. Of 159 patients with tuberculosis, 31 (19.5%) were unreported. A case was defined according to the Centers for Disease Control and Prevention definition. Unreported cases were less likely than previously reported cases to have specimens that were culture positive for M. tuberculosis, 14 of 31 (45.2%) compared with 111 of 128 (86.7%). Excluding the laboratory, tuberculosis diagnoses in acquired immunodeficiency syndrome registry patients had the highest predictive value of finding tuberculosis (94.1%), followed by tuberculosis clinic records (71.7%), and pharmacy listings (45.6%). Tuberculosis discharge diagnoses, however, yielded the largest number of unreported cases (14). Health care providers should be educated regarding the importance of promptly reporting all suspected TB cases regardless of results of laboratory testing.