The state of São Paulo is the Brazilian leader in fruit production. One of the main crops is guava, which currently occupies an area of 6,634 hectares with an annual production of approximately 200 thousand tons, demonstrating the growth and economic importance associated with both the high consumption of fresh fruit and products from its industrialization. The agro-industrial wastes from the processing of guava are equivalent to 10-15% of the total mass of the fruit, consisting of peel, macerated pulp and seeds. However, they are irregularly disposed of in landfills, which can contribute to contaminating the environment and wasting nutrients. These wastes are composed of several compounds among them, the fatty acids. In this context, one of the main FA found in agro-waste is linoleic acid (LA), which has attracted the attention of the scientific community due to its rich anticancer, antiobesity and antidiabetic properties, which makes it a promising food supplement. Thus, the literature reports several analytical methods for the determination of LA based on chromatographic, spectroscopic and spectrophotometric techniques. On the other hand, electrochemical sensors stand out for bringing together excellent advantages for analytical determinations, such as: portability, sensitivity, selectivity, low detection and quantification limits and relatively low instrument cost. However, this work reports for the first time the determination of LA in guava agro-waste on an rGO-modified surface decorated with FeNPs coated with molecularly imprinted poly(aniline). For the construction of the electrochemical platform, a glassy carbon electrode (GCE) was modified with rGO using a suspension of graphene oxide 0.50 mg mL^-1 in Na₂SO₄ 0.10 mol L^-1, where a potential of –1.4 V was applied until reaching 1 mC. Then, a 5.0×10^-3 mol L^-1 solution of FeCl₃ in KCl 0.10 mol L^-1 was used to carry out the electrodeposition of FeNPs on the GCE/rGO. The formation of FeNPs on the electrode surface occurred by applying a constant potential of –1.4 V until reaching an electrodeposition charge of 30 mC. The molecularly imprinted film was prepared from a 0.1 mol L^-1 aniline solution containing 3.0×10^-6 mol L^-1 LA in 0.5 mol L^-1 H₂SO₄. The electropolymerization process was carried out using cyclic voltammetry, where 15 voltammetric cycles were applied in a potential range of –0.2 to 1.0 V (50 mV s^-1). Subsequently, the electrode was characterized by scanning electron microscopy, X-ray energy spectroscopy, electrochemical impedance spectroscopy and cyclic voltammetry. The experimental factors that affect the analytical performance of the proposed sensor were optimized: monomer concentration, number of cycles for electropolymerization, extraction time and rebinding time. Using differential pulse voltammetry, the electrode exhibited a wide concentration range from 1.0×10^-12 mol L^-1 to 1.0×10^-10 mol L^-1, with a detection limit of 3.0×10^-13 mol L^-1 and excellent sensitivity (3.4×10⁷ L A mol L^-1). Sensor response was evaluated by intra-day and inter-day analyses. For intra-day analysis, a single sensor was used to carry out 20 successive detections of LA. In this study, the RSD value obtained was 1.2% (n = 20). For inter-day analysis, 15 electrodes were prepared under the same conditions and used separately for the detection of LA. The RSD value obtained was 4.9% (n = 15). The stability of the device was analyzed using an electrode prepared and stored in a solution at 5.0×10^-3 mol L^-1 [Fe(CN)₆]^3-/4- containing 0.1 mol L^-1 KCl in solution PBS buffer (0.1 mol L^-1, pH 7.0) at room temperature for 10 days. During this period, 5 consecutive measurements were taken alternately for LA detection. After 10 days, the electrode presented 90.1% of its initial current, demonstrating high stability for LA detection. The selectivity evaluation was performed by comparing the current intensity obtained in the reconnection step involving the application of the concentration of LA and the concentrations of other potentially interfering molecules. The results obtained show a high capacity of the proposed sensor for the selective recognition of LA. The determination of LA in the sample was carried out from the addition of standard. The sample was enriched with known concentrations of LA in the range 1.0×10⁻¹² mol L^-1 to 5.0×10 ⁻¹² mol L^-1. The concentration of LA found in a sample of guava agro-waste was (1.3 ± 0.1) mol L^-1 (n = 3). These results were validated through recovery trials, where the recovery rates obtained ranged from 93.5% to 111%, with RSDs ≤ 5%. The results obtained by the proposed sensor indicate that it can be successfully applied in the detection of LA in guava agro-waste.