To investigate the validity and reliability of the Portuguese version of the Test of Infant Motor Performance (TIMP) for Brazilian infants. The study was conducted with 20 health professionals and 655 Brazilian infants (between 34 weeks of Post-Menstrual Age and 17 weeks Corrected Age). (1) Content validity: High concordance among the experts, for language clarity and pertinence, were found (ICC from.74 to .100; AC1 from .87 to 1.00); (2) Inter-rater, intra-rater, and test-retest reliability: Strong inter ( from .65 to .99) and intra ( from .97 to .99) raters agreement, and high test-retest reliability ( = 1.00; ICC = 1.00); (3) Internal consistency: Excellent internal consistency ( from .71 to .98); (4) Discriminant validity: Significant power to identify groups at-risk for delays ( < .0001); (5) Predictive validity: Capability to predict motor delay ( .55 to .89) was found; and, (6) Concurrent validity: Significant ( < .05) and low correlations at the first months ( .21 to .24) and weak to moderate correlations at 3 and 4 month-old ( .42 and .44, respectively) were observed between the TIMP and Alberta Infant Motor Scale. The results provide evidence of validity and reliability of the Portuguese version of the TIMP for Brazilian infants.