There were two peoples that inhabited the Valle de Arán and the area which, much later, would become Andorra. They were the Arenosios and the Andocinos, quasi-Celtic pre-Roman urnfield cultures with influences from the Iberian Ilergetae people to the South and East. It is difficult to say exactly when this remote area came under Roman influence, but the coastal stretch to the East north of the Ebro became friendly to Roman interests as early as 218 BCE, when Cornelius Scipio Calvus landed at Emporion. Through a combination of diplomacy and being nice, he gained the support not only of the coastal region but also at least some of the inland mountain peoples, who had technically been under the control of the Carthaginian commander Hasdrubal:

While these things were happening in Italia, Gnaeus Cornelius Scipio [Calvus], having been sent with a fleet and an army to Spain, began from the mouth of the Rhone and sailed down the Eastern flank of the Pyrenees, landing finally at Emporiae. He disembarked his forces and, starting with the Laeetani, he brought the whole of the coastal peoples as far as the Ebro River within the sphere of Roman influence by renewing former alliances and forming new ones. By this he acquired a reputation for clemency which extended not only to the coastal peoples but to the more warlike tribes in the interior and the mountain districts. He established peace with these tribes as well, and what's more, he secured their support in arms and some strong auxiliary cohorts were enrolled from them (validaeque aliquot auxiliorum cohortes ex iis conscriptae sunt)(Livy 21.60).

This particular region was probably included in the reorganization of Spain in 198 BCE by the Senate, when the amount of Roman praetors was increased from 4 to 6 so that two praetors could govern the two Spains (Hispania Citerior and Ulterior). I can't find any concrete information on this little patch of the Pyrenees, though, and I could be wrong about the timing. After 199, the Romans pulled back the legions from Spain (inexplicably) and left only Italian auxiliaries there. Subsequently there was a large rebellion, but it doesn't seem as though the Arenosios or the Andocinos participated. When Cato arrived in the winter of 196/5 BCE to the north of the province, at Emporion, Livy says that only the Ilergetae were still loyal (33.21). I tend to doubt he meant the mountain tribes, though. Also, we should probably assume that Livy was drawing on Cato's first-hand reports when writing his history, and that a lot of Cato's reports were not exactly humble, reliable accounts.

After Cato's campaign, Roman attention moved westward and I can find no further mention of the little region of Andorra. It was probably already a nominal part of Hispania Citerior by 188 BCE, but again I'm not 100% sure on that and would welcome a correction.