O objetivo deste estudo foi analisar por meio de microscopia ótica e microscopia eletrônica de varredura a adaptação marginal cervical e a porosidade superficial de restaurações mésio-oclusal ou disto-oclusal em resina composta posterior obtidos in vitro em razão de três diferentes técnicas restauradoras: T1 = matriz de aço associada a cunha de madeira; T1 = matriz de poliéster associada a cunha de madeira e T1 = matriz de poliéster associada a cunha refletiva. Foram utilizados pré-molares superiores elou inferiores indicados para a exodontia por razões ortodônticas. Obtidas as restaurações e decorridos 7 dias, os dentes foram extraídos e submetidos a análise em microscopia ótica e microscopia eletrônica de varredura. Nestas condições, os autores concluem. 1 diferentes técnicas levam a diferentes valores de fenda marginal; 2. a técnica T1 apresentou maior fenda marginal (31,766 um), seguida pelas técnicas T1 (21,968 um) T, (23,569 um) iguais estatisticamente entre si. A porosidade superficial esteve presente em 80% das restaurações, independentemente da técnica utilizada.

The aim of this paper was to analyse, through optical and scanning electron microscope, the cervical margnal adaptation of MO or DO posterior composite resin restorations obtained in vivo by three different restorative techniques: T1 = metal matrix band associated to a wooden wedge; T2 = transparent matrix strip associated to a wooden wedge; T, = transparent matrix strip associated to a light - reflecting wedge. It was selected first mandibular and/or maxillary premolars, indicated for extraction. Modified Class II cavities were prepared, followed by restorative procedure. The same steps were followed in sequence throughout the three restorative techniques, only the matrix and wedge selection were varied. After 7 days, the teeth were extracted. The results of cervical adaptation were obtained from the gap extension measure in three different regions, in photomicrographics obtained by SEM. The superficial porosity was analysed in stereoscopic lupe ZEISS. The cervical gap measures, was observed, through a varying analysis, significant variability for the factor "technique". Concerning the factor "region" and the interaction "technique x region" we observed that it was non-significant. Therefore, we can conclude that: I. different techniques showed different values of marginal gap and independent of region analysed; 2. the T, technique showed bigger cervical gap (31.766 pm) followed by techniques, T, (21.968 pm) and T, (12.569 pm) and statistically equal; 3. superficial porosity is always present in 80% of the restorations.