There is a strong desire to safeguard taro (Colocasia esculenta) continued production by breeding for improved production. The two major constraints to the development of taro cultivation, processing and value-adding are: its insufficient adaptation to the changing climate (especially its limited adaptation to drought) and its chemotype (low dry matter, starch and amylose contents and high oxalate content). Both can be genetically improved. The chances of getting a high yielding clone, with excellent eating quality, are generally very low and they become much lower when the selection procedure includes additional traits. Therefore breeders have to test a lot of offspring individuals, even when the parents are the best existing genotypes. The sustainable and safe use of taro germplasm cannot be addressed if only one gene pool is studied. There is an urgent need to exchange genetic resources. This makes sense both in scientific terms and in terms of technical co-operation between developing countries. The present paper highlights some of the present constraints to taro genetic improvement and discuss future perspectives opened by the international exchange of true seeds.