A comparison has been made of the electrocrystallization of tantalum and niobium from baths containing either KF and K2TaF7 or KF and K2NbFT. The main purpose of the work was to determine why current efficiencies are generally much lower for niobium than for tantalum under the same experimental conditions. The phase diagram of the system KF-K2NbF7 has been drawn. Comparison with the diagrams of the system KF-K2TaF7 found in the literature does not give any evidence for a basic difference in the nature of the ionic species present in both types of baths. The results of the electrolyses performed have led to the following main conclusions: (A) An increase of the K2NbF7 or K2TaF7 content of the bath results in a considerable decrease of the current efficiency. This effect is much more striking for niobium than for tantalum. The grain size decreases also when the K2NbF7 or K2TaF7 content of the bath increases. (B) The current efficiency increases roughly with the cathodic current density while the grain size shows a very complex behavior against the same parameter. (C) The current efficiency decreases and the grain size increases when the length of the electrolysis increases. (D) The grain size decreases regularly when the temperature increases, while the current efficiency passes through a maximum at about 900~ Cathodic polarization curves drawn in both types of bath have shown a rather complex behavior of the ionic species to be discharged at the cathode. It appears that some part of the metallic deposit at the cathode is removed from it by the agitation of the bath and by its own weight and is redissolved in the bath through a chemical reaction giving rise to lower valency species. A part of these species is reoxidized at the anode, and another part of them participates again in the cathodic process. This phenomenon is much more effective for niobium than for tantalum in the same type of bath. All the experimental observations are interpreted in terms of this mechanism.It is well known that tantalum and niobium have very similar chemical properties, so similar indeed that it is necessary to use very elaborate methods to separate them during the mineral and chemical processing of their ores.It is thus quite astonishing that, while tantalum can be prepared economically by fused salt electrolysis, all the attempts to prepare niobium in the same way failed, the current efficiency being very low. It should of course be very interesting to know why there is such a big difference in the behavior of the two metals during the electrolysis. In this respect, although the literature (1) contains a rather large number of papers on the electrowinning of tantalum and niobium, it is rather difficult to draw any general conclusion from the published results. This is mainly due to the fact that most of the researchers have worked only on one of the two metals and generally under conditions so different that any valuable comparison of the results is impossible. It should be pointed out however that Chuk-Ching Ma (2) has studied the ...