Background: In the last decade, an innovative approach has emerged for arthropod identification based on matrixassisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS). Increasing interest in applying the original technique for arthropod identification has led to the development of a variety of procedures for sample preparation and selection of body parts, among others. However, the absence of a consensual strategy hampers direct inter-study comparisons. Moreover, these different procedures are confusing to new users. Establishing optimized procedures and standardized protocols for mosquito identification by MALDI-TOF MS is therefore a necessity, and would notably enable the sharing of reference MS databases. Here, we assess the optimal conditions for mosquito identification using MALDI-TOF MS profiling. Methods: Three homogenization methods, two of which were manual and one automatic, were used on three distinct body parts (legs, thorax, head) of two mosquito laboratory strains, Anopheles coluzzii and Aedes aegypti, and the results evaluated. The reproducibility of MS profiles, identification rate with relevant scores and the suitability of procedures for high-throughput analyses were the main criteria for establishing optimized guidelines. Additionally, the consequences of blood-feeding and geographical origin were evaluated using both laboratory strains and fieldcollected mosquitoes. Results: Relevant score values for mosquito identification were obtained for all the three body parts assayed using MALDI-TOF MS profiling; however, the thorax and legs were the most suitable specimens, independently of homogenization method or species. Although the manual homogenization methods were associated with a high rate of identification on the three body parts, this homogenization mode is not adaptable to the processing of a large number of samples. Therefore, the automatic homogenization procedure was selected as the reference homogenization method. Blood-feeding status did not hamper the identification of mosquito species, despite the presence of MS peaks from original blood in the MS profiles of the three body parts tested from both species. Finally, a significant improvement in identification scores was obtained for field-collected specimens when MS spectra of species from the same geographical area were added to the database. Conclusion: The results of the current study establish guidelines for the selection of mosquito anatomic parts and modality of sample preparation (e.g. homogenization) for future specimen identification by MALDI-TOF MS profiling. These standardized operational protocols could be used as references for creating an international MS database.