The technology transfer from technological institutes to enterprises is a subject of increasing interest for the development of Innovation National Systems. For the purpose of fostering technology transfer, it becomes crucial to identify those technological institutes of relatively high performance. Likewise, it is interesting to show the characteristics of those organizations as well as the way by which they use inputs to produce the expected outputs. The DEA model constructs an efficient frontier with those observations of relatively best perfomance; it shows for each inefficient unit what are the reasons behind such a low perfomance; and finally, it determines efficiency targets for each inefficient unit. This article seeks to identify the efficiency frontier from a sample of Spanish and Brazilian technological institutes and what may be the paths to improve the current performance. Our results inform about the standard characteristics of a technological institute operating in the frontier; the priorities among production targets: and the great difference in efficiency levels within the sample. Our conclusions show that there may be a significant scope for improvement and the need of practice review in the process of technological transfer.La transferencia y comercialización de tecnología desde los Institutos Tecnológicos (ITs) a las empresas es un tema de gran interés para el desarrollo de los sistemas nacionales de innovación. La identificación de aquellos institutos que consiguen la máxima eficiencia en su papel de transferir tecnología a las empresas, las caracteristicas de estos institutos y cómo utilizan sus inputs (ej. presupuestos, investigadores, recursos para transferencia tecnológica, dotación de personal administrativo, etc.) para la consecución de unos outputs (ej. número de contratos de transferencia tecnológica, número de publicaciones, etc), resulta interesante cuando se desea impulsar el proceso de transferencia y comercialización de tecnologia de un sector, una región o un país. La evaluación DEA construye una frontera de eficiencia con aquellas observaciones de mayor rendimiento relativo. La técnica de “Análisis Envolvente de Datos” (DEA) localiza en la curva-frontera únicamente los ITs que exhiben los niveles más altos de productividad; muestra para cada IT juzgado como ineficiente las razones de ese status; y finalmente, produce una serie de objetivos que permitirían reconducir las situaciones de bajo rendimiento. El artículo pretende mostrar una visión clara de cuales son los ITs españoles y brasileños considerados eficientes y que actuaciones pueden llevarse a cabo para invertir la situación y reducir las bolsas de ineficiencia identificadas en el sector. Los resultados informan de las características de ITs representativos de la eficiencia técnica e identifican prioridades alternativas entre objetivos. Asimismo, los resultados muestran una amplia heterogeneidad en niveles de eficiencia entre ITs y por consiguiente, un significativo ámbito de mejora y de revisión de actividades.The technology transfer from technological institutes to enterprises is a subject of increasing interest for the development of Innovation National Systems. For the purpose of fostering technology transfer, it becomes crucial to identify those technological institutes of relatively high performance. Likewise, it is interesting to show the characteristics of those organizations as well as the way by which they use inputs to produce the expected outputs. The DEA model constructs an efficient frontier with those observations of relatively best perfomance; it shows for each inefficient unit what are the reasons behind such a low perfomance; and finally, it determines efficiency targets for each inefficient unit. This article seeks to identify the efficiency frontier from a sample of Spanish and Brazilian technological institutes and what may be the paths to improve the current performance. Our results inform about the standard characteristics of a technological institute operating in the frontier; the priorities among production targets: and the great difference in efficiency levels within the sample. Our conclusions show that there may be a significant scope for improvement and the need of practice review in the process of technological transfer.