Abstract Lipoarabinomannan (LAM), a cell wall component of mycobacteria, can be detected in the urine of tuberculosis (TB) patients. Advantages of this diagnostic include the ease of sample collection and test methods. However, as with most new TB diagnostics, LAM tests have been evaluated in well-controlled laboratory settings and subsequently need assessment under real working conditions. Our experience showed that the diagnosis of TB using the detection of LAM in urine under field conditions is prone to false-positive results due to contamination. Dust and soil, but also stool, seemed to lead to increased OD values and thus false-positive results of the enzyme-linked immunosorbent assay (ELISA) for LAM; however, contamination with blood, as well as bacterial or fungal organisms, had no influence. The collection of urine for the detection of LAM should therefore follow strict collection criteria in order to avoid contamination.