Abstract Background Invasive mold diseases (IMDs) cause severe illness, but public health surveillance data are lacking. We describe data collected from a laboratory-based, pilot IMD surveillance system. Methods During 2017–2019, the Emerging Infections Program conducted active IMD surveillance at 3 Atlanta-area hospitals. We ascertained potential cases by reviewing histopathology, culture, and Aspergillus galactomannan results and classified patients as having an IMD case (based on European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group [MSG] criteria) or a non-MSG IMD case (based on the treating clinician’s diagnosis and use of mold-active antifungal therapy). We described patient features and compared patients with MSG vs non-MSG IMD cases. Results Among 304 patients with potential IMD, 104 (34.2%) met an IMD case definition (41 MSG, 63 non-MSG). The most common IMD types were invasive aspergillosis (n = 66 [63.5%]), mucormycosis (n = 8 [7.7%]), and fusariosis (n = 4 [3.8%]); the most frequently affected body sites were pulmonary (n = 66 [63.5%]), otorhinolaryngologic (n = 17 [16.3%]), and cutaneous/deep tissue (n = 9 [8.7%]). Forty-five (43.3%) IMD patients received intensive care unit–level care, and 90-day all-cause mortality was 32.7%; these outcomes did not differ significantly between MSG and non-MSG IMD patients. Conclusions IMD patients had high mortality rates and a variety of clinical presentations. Comprehensive IMD surveillance is needed to assess emerging trends, and strict application of MSG criteria for surveillance might exclude over one-half of clinically significant IMD cases.