Mobile herbicides have a high potential for contaminating groundwater. Biochar have been proposed as an alternative to retain herbicides and modify soil microbial community, thus, limiting the transport of herbicides and other contaminants in soil and water sources. The objective of this research was to evaluate the effect biochar amendments on sorption-desorption and mineralization of mobile herbicides in agricultural soils. In chapter one, sorption-desorption were evaluated using the batch equilibrium method at five concentrations of hexazinone, metribuzin and quinclorac. Soil was amended with eucalyptus (Eucalyptus grandis), rice hull (Oryza sativa), and native bamboo (Merostachys skorvotzii) biochar at rate of 0 (control – unamended) and 1% (w w-1), corresponding to 0 and 12 t ha-1, respectively. The highest sorption in unamended soil followed the decreasing order of herbicides: quinclorac (65.9 %) > metribuzin (21.4 %) > hexazinone (16.0 %). Native bamboo biochar provided the highest sorption in comparison to rice hull and eucalyptus biochar amended soils for the three herbicides. The highest desorption in unamended soil followed the decreasing order of herbicides: metribuzin (18.35 %) > hexazinone (15.9 %) > quinclorac (15.1 %). Addition of native bamboo biochar provided the lowest desorption among the biochar amendments on the three herbicides. In chapter two, sorption and desorption were evaluated using the batch equilibrium method at five concentrations of metribuzin. Three soils were amended with grape wood (Vitis vinifera L.) and loblolly pine (Pinus taeda) biochars at rate of 0 (control – unamended) and 5% (w w-1), corresponding to 0 and 60 t ha-1, respectively. Mineralization of 14C-metribuzin was performed in biometer flasks at a rate of 4 mg·kg−1 (target concentration). The 14CO2-metribuzin released from mineralization was trapped in 0.4 M sodium hydroxide solution (NaOH) analyzed at each sampling time (1, 3, 7, 14, 21, 28, 35 and 42 days). Sorption of metribuzin on the three unamended soils followed the decreasing order: drummer (61.3 %) > oakville (24.56 %) > ipava (22.35 %). Grape wood biochar (82.6 – 83.3 %) showed the highest amount sorbed compared to loblolly pine biochar (43.9 – 68.4 %) in all the three soils. Ipava soil provided the highest amount of desorbed metribuzin among the unamended soils in both, first (40.52%) and second (57.97%) desorption step. Loblolly pine biochar provided the lowest amount desorbed on first (16.4 – 29.5 %) and second (24.7 – 42.5 %) desorption steps between the two biochars in all the three soils. The two consecutive steps of metribuzin desorption showed that more than 65% of the sorbed metribuzin was retained in the soils amended with grape wood biochar. Metribuzin mineralization in drummer, ipava and okaville soil was low in comparison to the initial concentration applied. The addition of grape wood and loblolly biochar decreased mineralization and increased the non-extractable amount of metribuzin, but this effect was only observed for some soils. As conclusion, the addition of biochar is a good alternative to increase the sorption and avoid the transport of mobile herbicides in agricultural soils.