The informal singular irregular imperatives of Spanish have traditionally been explained as the result of primarily phonological processes. However, such an approach presents various problems which to date are unresolved: d?c > di, but f?c > faz > haz, not *fa > *ha. v?n? > ven, without diphthongization because of ?, but t?n? > ten, not *tien. Ten has been explained as analogical to ven, but a form like *vien could just as well have resulted from analogy with the phonologically expected *tien. Also, s?d? > sé, without expected diphthongization, a form that cannot be explained as analogical to ven. Finally, pon, sal, ten, ven, (val) have been explained as resulting from apocope during the Middle Ages, but the process is not attested in the imperative (i.e., *pone > *pon(e) > pon, etc.) as it is in the present indicative (e.g., pone > pon(e) > pon). Furthermore, it has not been satisfactorily explained why apocope was permanent in these imperatives but not in the present indicative (and other tenses). In this study, all irregularities, synchronic and diachronic, are explained through a new analysis based primarily on a process of historical morphology, namely, back-formation.