A student from Guinea-Bissau presented with a 2-year history of painful swollen lymph nodes in the right side of her neck. Some had opened and formed fistulas that suppurated (Figure 1A). Smears of pus were stained with auramine O (AuO) and screened for mycobacteria at 200 magnification. Very rare images with the morphology of acid-fast organisms were seen in two samples. Following current recommendations, the AuO-stained smears were restained with Ziehl–Neelsen (ZN) to confirm the findings of these images.However, threeexperiencedmicroscopistsobservednoacidfast bacteria, and the ZN-stained smears were reported as negative. Later, Mycobacterium tuberculosis complex grew from all samples. Because ZN-stained mycobacteria fluoresce red–orange when excited with green light, the ZN-stained smears were reexamined with a fluorescence microscope (excitation 546 nm, emission >590 nm). Comparing the same microscopic field, the mycobacteria invisible in transmitted light (Figure 1, top panel: ZN) were clearly visible by fluorescence (Figure 1, bottom panel: ZN-Fl). Because AuO is not excited at 546 nm, smears that are only stained with AuO show no red fluorescence when examined with the filter combination ‘excitation: 546 nm/emission >590 nm’. This case illustrates a disturbing aspect of ZN-stained smears for the diagnosis of tuberculosis. It appears that fluorescence microscopy allows visualization of more mycobacteria, using either AuO or ZN, than does transmitted light microscopy of ZNstained smears. In paucibacillary specimens ZN-stained smears alone may yield false-negative results.