This paper presents an extraction method based on dispersive-nanoparticle-solid phase microextraction (DNSPME) for the preliminary preconcentration and subsequent spectrophotometric determination of trace amounts of malachite green (MG). The application of nanoparticles permits the easy separation and extraction of MG from trout fish water and natural water samples. The analyte was accumulated on a gamma-Fe2O3 nanomaterial loaded on activated carbon (gamma-Fe2O3-NPs-AC) that was identified by FESEM, XRD, FTIR, EDS and UV-Vis techniques. The influence of the expectable parameters on the extraction recovery, according to p < 0.05, was investigated and judged using a two-level Plackett-Burman screening design with 7 variables (adsorbent mass, centrifugation time, eluent volume, ionic strength, pH, ultrasonication temperature and ultrasonication time). It was found that three significant variables namely adsorbent mass, eluent volume and pH have a great influence on optimization using a central composite design combined with a desirability function. The results showed that the semi-empirically obtained second-order model was able to efficiently predict the ER% for MG adequately with a coefficient of determination of 99.7% (p < 0.001); the higher efficiency of the model was obtained through a good compromise between the experimental and predicted data. Working under the optimum conditions, specified as 0.6 mg of adsorbent and 120 mu L of eluent volume at pH 6.0, led to the achievement of a high and reasonable linear range over 1-4000 ng mL(-1) (R-2 = 0.999) with a detection limit of 0.175 ng mL(-1) and an obtained quantification limit of 0.583 ng mL(-1). The relative standard deviation (RSD) for ten replicates was less than 3.50%. The proposed method was successfully applied for the determination of MG in trout fish water and natural water samples with excellent recoveries corresponding to spiked samples. All these results prove the suitability of the present method in terms of simplicity, easy operation conditions, efficiency and sensitivity for the determination of MG in real samples.