In this study, we investigate the impact of the share of the foreign labor force on the wage of native workers in Portugal between 2010 and 2019 using linked employer-employee data from Quadros de Pessoal. By leveraging job characteristics from the O*NET skill taxonomy, we create more homogeneous skill groups, enabling a precise analysis of immigration's impact on specific skill sets. The empirical analysis, focusing on occupation-experience groups, reveals a positive association between native wages and immigrant shares. In contrast, when groups are based on education-experience, the relationship appears negative. These contradictory findings suggest that the impact of immigration on native wages varies significantly depending on how labor markets are segmented. Furthermore, our analysis demonstrates a positive and statistically significant effect on native wages in high-skilled occupations, while native wages in low-skilled occupations are negatively affected due to increased competition. Our findings highlight the importance of considering occupation classification over simple education levels and suggest that diverse results in existing literature may be due to sample averaging.