Net profit, annual work volume, and working capital can be considered as the main financial performance indicators for any construction company. Sufficient liquidity must be properly assessed to ensure the survival of the business in both short-term and long-term bases. Large amount of working capital simply means idle funds in a form of current assets that does not gain any profit for the company. On other hand, small amount of working capital means that the company is unable to meet its liabilities and it faces complexity to participate in new project tenders, as a consequence its annual work volume might be decreased. Then, the excess or shortage of working capital affects badly the companies' profitability. Hence, it is obvious that the construction companies' working capital, net profit, and annual work volume constitute three interrelated financial performance indicators that have to be appropriately assessed. The present study aims to develop a model to help the construction companies' managers to assess and forecast their companies' financial performance indicators: working capital, net profit, and annual work volume. Through this research, the genetic algorithm technique (GA) will be integrated with the neural network technique (NN) to develop the proposed model. The developed model will be able to predict the three financial performance indicators: working capital, net profit, and annual work volume, for an upcoming year based on previously published financial statements data. A comprehensive literature review was conducted and 23 factors were identified as the most influencing factors on the construction companies' financial indicators: working capital, net profit, and annual work volume. One hundred and sixty four Egyptian construction companies' financial statements were gathered and analyzed to extract data regarding the identified 23 factors. The extracted data were used to develop a NN-GA hybrid and NN only models to assess the construction companies' financial indicators. The two developed model outputs are compared to evaluate their predictive capability. This comparison showed that, the NN-GA hybrid model predictive capability is better than the NN only model predictive capability. Incorporating the GA enhances the predicting capability of the developed model by an average of 4.0%. Key words: financial indicators, construction company's working capital, firm profitability, annual work volume, genetic algorithm, neural network. On peut considerer le benefice net, le volume annuel de travail et le fonds de roulement comme les principaux comme les principaux indicateurs de performance financiere de n'importe quelle entreprise de construction. Les liquidites doivent etre suffisantes et correctement evaluees pour assurer la survie de l'entreprise a court et long termes. D'importants fonds de roulement impliquent tout simplement des fonds libres sous forme d'actifs circulants qui ne rapportent aucun profit a l'entreprise. Par ailleurs, peu de fonds de roulement signifient que l'entreprise est incapable d'honorer ses dettes et a des difficultes a s'impliquer dans de nouveaux projets, consequence de la diminution du volume de travail. Par la suite, l'exces ou le manque de volume de travail affecte negativement la rentabilite de l'entreprise. Ainsi, il est evident que, dans le cas des entreprises de constructions, le fonds de roulement, le benefice net et le volume de travail annuel constituent trois indicateurs de performance financiere interrelies devant etre correctement evalues. La presente etude a pour but de developper un modele visant a aider les dirigeants des entreprises de construction a evaluer et prevoir les indicateurs de performance financiere de leur societe : fond de roulement, benefice net et volume de travail annuel. Par l'intermediaire de cette etude, la technique des algorithmes genetiques (AG) sera integree a la technique des reseaux de neurones (RN) afin d'elaborer le modele propose. Ce dernier pourra predire les trois indicateurs de performance financiere--fonds de roulement, benefice net et volume de travail annuel--une annee a l'avance, a partir des donnees des etats financiers deja publies. Une analyse approfondie de la litterature a ete menee et 23 facteurs ont ete identifies comme les facteurs qui influent le plus sur les indicateurs de performance financiere des entreprises de construction (fonds de roulement, benefice net et volume de travail annuel). Cent soixante-quatre etats financiers d'entreprises de construction egyptiennes ont ete rassembles et analyses de maniere a en extraire les donnees relatives aux 23 facteurs identifies. Ces donnees ont ete utilisees pour l'elaboration de modeles hybride RN-AG et RN seul afin d'evaluer les indicateurs financiers des entreprises de construction. Les resultats des deux modeles developpes ont ete compares pour evaluer leur capacite predictive. Cette comparaison a montre que la capacite predictive du modele hybride RN-AG est meilleure que celle du modele RN seul. L'integration de la technique des AG ameliore la capacite predictive du modele developpe de 4% en moyenne. [Traduit par la Redaction] Mots-cles: indicateurs financiers, fonds de roulement d'une entreprise de construction, rentabilite de l'entreprise, volume annuel de travail, algorithmes genetiques, reseau de neurons., Introduction Company's capital requirements can be generally divided into fixed capital and working capital. The fixed capital represents the capital needed to cover the long-term financial requirements. Working capital represents [...]