Invasive aspergillosis (IA) is a disease with a complex diagnosis due to its non-specific clinical manifestations, which delays treatment and consequently leads to a worsening of the disease. The Aspergillus galactomannan (GM) lateral flow assay was evaluated in pretreated sputum (PS) and serum samples for the diagnosis of IA. Clinic epidemiological data were collected from 76 patients at the Instituto Médico La Floresta between January 1, 2022, and April 30, 2024. The GM lateral flow assay was performed on serum, BAL (bronchoalveolar lavage), and PS samples for subsequent classification according to the criteria established by the EORTC/MSG into possible, probable, proven IA, and no IA. Only 28 patients were presented with possible IA, 28 with probable IA, and 20 did not present with IA. A descriptive analysis was applied to 32 patients with galactomannan results ≥ 0.7 in serum and ≥ 0.8 in sputum or BAL, yielding statistically significant values for sputum and the associated serum (p=0.00), and for BAL (p=0.041) and its accompanying serum (p=0.05). Furthermore, the means obtained were within the 95% confidence intervals, being consistent and reliable. A correlated t-test was performed on 3 patients who provided all three sample types, revealing p (0.453) > 0.05, which indicates no significant difference between pretreated sputum and BAL. PS is an alternative sample to BAL for high-risk patients. The use of respiratory samples (BAL/PS) increases the probability of galactomannan detection in non-neutropenic patients compared to the use of serum and can be used for IA classification; however, it is considered that EORTC/MSG criteria have limited applicability in this population.