Chlorophyll fluorescence quenching induced by low concentrations of m-dinitrobenzene (DNB) is investigated. In intact spinach chloroplasts DNB causes photochemical and non-photochemical quenching. The two forms of quenching are distinguished by applying the saturation pulse method with a new type of modulation fluorometer. Half-maximal photochemical quenching is observed at about 3 micromolar DNB. It is inhibited by 3-(3,4 dichlorophenyl)-1, 1-dimethylurea (DCMU) and by 2,5-dibromo-3-methyl-6-isopropyl-p-benzoquinone (DBMIB). Photochemical quenching by DNB leads to suppression of the I-P transient in a fluorescence induction curve. Upon application of saturating continuous light, the increase of fluorescence yield is separated into a photochemical and a thermal part. DNB causes suppression of only the slowest sub-component of the thermal part, in analogy to the action of Hill reagents. Simultaneous measurements of oxygen exchange rate and fluorescence reveal that a part of DNB induced quenching is accompanied by oxygen uptake. Most DNB-induced non-photochemical quenching is prevented by nigericin and, hence, can be considered "energy-dependent" quenching. The small component persisting in the presence of nigericin is identical to the one observed with methylviologen and other Hill reagents, likely to be due to static quenching by oxidized plastoquinone. The presented data confirm the original finding of Etienne and Lavergne (Biochim Biophys Acta 283: 268-278, 1972) that low concentrations of DNB selectively affect the thermal component of variable fluorescence. However, while these authors interpreted the quenching by a non-photochemical mechanism, the present investigation emphasizes a photochemical mechanism, in analogy to the effect of electron acceptors or mediators.