L’objectif de ce travail est de réaliser une caractérisation complète des margines brutes et décantées et d’étudier la diminution de la charge organique et des polyphénols en utilisant les techniques de coagulation-floculation.Les essais de coagulation ont été réalisés à l’aide d’un banc de jar-test, constitué d’une série de six béchers. La série comporte une suspension témoin sans addition de coagulant ainsi que la même suspension soumise à des doses croissantes du coagulant (la chaux seule, le sulfate d’aluminium seul et la chaux combinée avec le sulfate d’aluminium).Les tests de coagulation-floculation montrent que l’application de sulfate d’aluminium à une dose de 1,5 g/L et à un pH entre 6,31-7,08 permet d’éliminer 40 % de la DCO, 27 % des MES et 41 % des polyphénols, tandis que l’application de la chaux à une dose de 20 g/L permet d’éliminer 43 %, 75 % et 50 % de la DCO, des polyphénols et des MES respectivement. La combinaison d’une dose de 1,5 g/L de sulfate d’aluminium et de 20 g/L de chaux permet de réduire 70,5 % des MES, 38 % de la DCO, 54 % des polyphénols et 61 % de la coloration.Il ressort de l’ensemble des résultats que le meilleur traitement par coagulation-floculation est obtenu par l’application de la chaux seule à une dose de 20 g/L et par la combinaison de 1,5 g/L de sulfate d’aluminium et de20 g/L de chaux., Olive mill wastewaters (OMW) are a significant source of environmental pollution, especially in important olive oil producing countries such as Spain, Italy, Greece, Tunisia, Morocco, Turkey, Lebanon, Syria and Portugal. When discharged into the environment, olive mill wastewaters create serious environmental problems, such as colouring of natural waters, alteration of soil quality, phytotoxicity and nuisance odours.Several methods have been reported for the removal of pollutants from these effluents. These technologies can be divided into three categories: biological, chemical and physical. Physical-chemical treatment is well known and has yielded promising results. Several authors have tested coagulation-flocculation techniques using different coagulants, such as aluminium sulphate, ferric chloride, ferric sulphate and lime, but there are few studies that have investigated the combined effect of a mixture of coagulants.The objectives of this work were to carry out a complete characterization of the raw and decanted olive mill wastewater, resulting from a modern unit located in the Marrakesh region, and to study the reduction in organic load and phenol content achievable by coagulation-flocculation using two different coagulants (lime, aluminium sulphate) and their combination. The quantity of sludge produced by the different coagulants was also determined.Coagulation tests were realized using jar test equipment in a series of six flasks. One flask corresponded to a control suspension without adding any coagulant. The other suspensions were treated with increasing coagulant concentrations of lime or aluminium sulphate, used separately (concentrations varying from 0 to 30 g/L and from 0 to 3 g/L, respectively) or in mixture. All solutions were stirred first for 3 min at 130 rpm, and then after the coagulant addition for 20 min at 30 rpm, followed by 1 h settling. Solution pH was adjusted to the desired value with sulphuric acid (4 N) or NaOH (5 N) before the coagulant was added. The supernatant was separated from the precipitate for analysis and the sludge was determined gravimetrically after drying at 100°C for 4h.Raw olive mill wastewaters are acidic (pH 4.55) and due to this low pH value biological treatment of OMW is limited. OMW are also highly saline (electric conductivity 8.4 mS/cm), due to the salting practiced to preserve olives during trituration. These effluents contain also high loads of Chemical Oxygen Demand (COD of 72 g/L) and toxic polyphenols (1.4 g/L). Comparison of the characteristics of the raw and decanted olive mill wastewater showed that decantation for 3 months did not seem to affect the physical-chemical characteristics of the studies olive mill effluent, except for the concentration of the total suspended solids (TSS), which passed from 57 to 32 g/L.Coagulation-flocculation tests showed that the application of aluminium sulphate without correction of the pH caused a small decrease of the pH from 4.67 to 4.54. The optimal elimination of the polyphenols (23%), TSS (24%) and colour (15%) was obtained with an amount of 1.5 g/L, whereas the optimum for elimination of COD (50%) was obtained with an amount of 1.8 g/L. These abatements demonstrate that the studied pH range did not allow very important elimination of the polluting loads, in particular for polyphenols. Adjust of the pH to 6.31-7.08 led to a slight improvement in the percentage of reduction of TSS (27%) and COD (40%), and the abatement of polyphenols passed from 23% to 41%. This application of aluminium sulphate alone generated only a small quantity of sludge, corresponding to the low degree of TSS removal. Lime application involved an increase in pH up to an addition of 15 g/L, after which the pH stabilized around 12. A lime application of 20 g/L resulted in the reduction of COD (43%), TSS (50%) and an important elimination of polyphenols (75%); sludge produced sludge after coagulation was significant (35 g/L).The successive addition of lime to 1.5 g/L of aluminium sulphate raised the pH, starting from an amount of 10 g/L, and induced a removal of TSS (71%), COD (38%), polyphenols (54%) and colour (61%) at a dose of 20 g/L of lime. The successive addition of aluminium sulphate to 15 g/L of lime involved only a slight variation of the pH and allowed the elimination of TSS (48%), COD (36%), polyphenols (35%) and colour (70%) at a dose of 3 g/L of aluminium sulphate. Comparison between the two coagulant combinations indicated that the best TSS removal was obtained for an amount of 1.5 g/L of aluminium sulphate and 20 g/L of lime, but with a production of 40 g/L of sludge. In the case of the elimination of colour, the best combination was obtained for a dose of 3 g/L aluminium sulphate and 15 g/L of lime, with a production of only 30 g/L of sludge. All in all, the best coagulation-flocculation was obtained by application of 20 g/L of lime (used alone) and by the combination of 1.5 g/L of aluminium sulphate and 20 g/L of lime, which gave a better elimination of the colloidal particles, a good reduction of the organic matter causing colour and the toxic polyphenols. The quantity of the sludge produced from the test with lime alone was 35 g/L, whereas the combination of lime and aluminium sulphate produced 40 g/L. Analysis of variance showed that the production of sludge for the two coagulants did not present statistically significant differences (p>0.05).