This article describes a pilot study on the metalinguistic abilities of 9 year-old Portuguese children as measured by the Portuguese version of a specific test for this age, the THAM-2 (Couceiro-Figueira, Pinto 2018), translated and adapted from the original Italian instrument (Pinto, Candilera & Iliceto 2003), already validated on a sample of 425 monolingual Italian-speaking children, recruited from various regions of Italy (age range : 9-14; grade 4 to 8). The test comprises six subtests (Comprenhension, Synonymy, Acceptability, Ambiguity, Grammatical function and Phonemic segmentation).Two types of questions are systematically posed in each subtest : a first, global question about a given issue, called Linguistic question (Lq), which measures an intuitive, epilinguistic level, and a second question, called metalinguistic (MLq) which requires that the previous answer be justified. To implement this justification, language must be analyzed by means of language itself, thus complying with a fully metalinguistic level, using intentional, focused, and explicit reasoning on language. The factorial structure of the original Italian test reflects these two levels of complexity of the questions/answers, which is also the case for the corresponding Spanish version, recently validated (Nunez Delgado, Pinto 2015). Twenty-seven children (mean age: 9,6; 16 F; 11 M), enrolled as fourth graders in a mid-size city of Portugal have been recruited. This pilot study presents descriptive and inferential basic statistics, subtest by subtest and in the total test. The results are discussed in view of a possible validation of the Portuguese THAM-2.