Thyroid pigments in black thyroid glands from minocycline-treated patients were compared by light and electron microscopy, histochemistry, and energy-dispersive x-ray analysis with minocycline-induced pigment in thyroid glands of laboratory animals, and with naturally occurring lipofuscins in untreated laboratory animals and humans. All thyroid samples examined contained nonbirefringent, Schmorl-positive pigment. However, the pigments in black thyroids from minocycline-treated patients resembled lipofuscins of untreated humans since both fluoresced and were Ziehl-Neelsen-and Sudan IV-positive. Minocycline induced pigment in rats was nonfluorescent and Ziehl-Neelsen-and Sudan IV-negative. Ultrastructurally, pigments in black thyroid glands of minocycline-treated humans resembled lipofuscins in untreated humans, and initial elemental analyses yielded similar spectra. Repeated analyses of the most electron-dense pigment deposits yielded spectra that resembled those of minocycline-induced pigment in laboratory animals-ie, both contained calcium. Black thyroid glands associated with minocycline administration contained predominantly lipofuscins with a small amount of another, possibly minocycline-related pigment. The absence of functional changes in patients and animals given minocycline suggests that discoloration of the thyroid gland associated with minocycline administration is innocuous. This is further supported by the lack of documented changes in thyroid physiology in patients that have received tetracyclines for a variety of indications in the last 30-odd years since their introduction to therapy.