This short paper addresses a very vexed issue, to which a huge literature has been dedicated so far: the origin of the so-called gerunds and gerundives in Latin. Any previous attempt has proved un- conclusive, mainly because of the proliferation of ad hoc rules assumed to account for the nd-forms and even more because of the plethora of solutions. Instead of assuming another etymon for the sake of antago- nism, this paper intends to reassess the whole issue within Latin itself: as shown by non-standard syntacti- cal features of Plautinian and Late Latin, there is a morphological relationship between the present parti- ciple and the ndō-gerund, used to express simultaneity. Whereas the previous scholarship has taken for granted the assumption that thematic verbs used to have a *-odno- suffix (cf. OLat. -und-), which led to tautological reconstructions totally unparalleled outside Italic, I would tentatively explain the unexpected o-grade of such forms by a crossing with the old o-grade participles (cf. OLat. *uoluns ‘willing’ reflected by uoluntās). Such an approach vindicates the ancient theory according to which -andus reflects *-ātan-ó- (< PIE *-eh2-t>mṇn-ó-), provided one assumes that a reanalysis of *-ātan-ó- was made as a “suffix” *-tanos following the thematic vowel of the first conjugation, which produced *fer-e-tnó- ‘ferendus’ from thematic *fer-e- ‘to bear’.