resumen Como bien se sabe, la caracterizacion del conocimiento en terminos de “creencia verdadera justificada” (CVJ) se ha considerado fallida desde la popularizacion de contraejemplos tipo Gettier. En este articulo se revisa el trabajo seminal de Gettier y sus argumentos. Se sostiene que los contraejemplos de Gettier se basan en movi -das de sustitucion ilegitimas, que uno de sus argumentos luce persuasivo porque confunde validez sintactica con verdad, que por tales razones su caso es mas debil de lo que aparenta, y que en realidad hay una salida para el partidario de la CVJ. Se sostiene que los casos tipo Gettier no constituyen contraejemplos genuinos para la caracterizacion del conocimiento en terminos de CVJ, y que, en consecuencia, dicha caracterizacion no resulta seriamente amenazada. Palabras claves: Gettier, creencia, conocimiento, razonador normal. AbstrAct As it is well known, the characterization of knowledge in terms of “Justified True Belief” (JTB) has been deemed unsuccessful since the popularization of Gettier-type counterexamples. This paper revisits Gettier’s seminal work and examines his arguments carefully. It holds that Gettier counterexamples are based on un-warranted substitution moves; that one of his arguments seems persuasive because it conflates syntactic validity with semantic truth; that for such reasons his case is weaker than it appears; and that there is, in fact, an avenue for escape open to the supporter of JTB. In short, I shall contend that Gettier’s cases are not genuine counterexamples to the standard characterization of knowledge in terms of JTB and that, consequently, such characterization is not seriously affected.