Louis II died in 875 without male heirs. He was succeeded by a series of short-lived uncles and cousins, who came from either France or Germany and stayed in Italy as short a time as possible. But after the fall of the last of these, Charles the Fat (king in Italy 879–887), most of the Carolingian kingdoms turned to non-Carolingian aristocratic families to rule them. In Italy, Berengar I, a female-line Carolingian and also marquess of the still-important border area of Friuli, was well placed to be elected as a king with genuine Italian commitments in 888. However, since Carolingians did not have a monopoly over the succession, anyone could claim the kingship; indeed, Berengar during his long reign (888–924) faced five such rivals, most of them militarily more successful than he was. Berengar was, in fact, not only long-lived but also unpopular; he spent much of the early part of his reign confined in his power base, Friuli. Even when he did not have internal rivals, as in 898–900, he was unlucky; in 899 the Hungarians invaded Italy, destroying Berengar’s army and initiating a series of raids that were to last, off and on, until the 950s.

Berengar I’s reign was a key period in Italian history. At its beginning the Italian kingdom was still a powerful and coherent institution, worth fighting civil wars to control. By his death the relevance of kingship itself was in doubt. This development resulted partly from Berengar’s personality, which was unadventurous and, militarily, unusually inept—but only partly. As the Carolingian political system had settled in, over four generations, local politics had become more stable and inward-looking. Hereditary families had taken over many counties, particularly the big marches of Friuli, Tuscany, and Spoleto. Sometimes local power was balanced between count and bishop, and the king’s capacity to intervene locally increasingly depended on their ability to maintain this balance of power. They usually accomplished this by supporting bishops, conceding more judicial and administrative power to them, particularly after 888. Sometimes, as at Bergamo or Cremona, counts were excluded from inside the city walls altogether. This was occasionally dangerous, for bishops, however loyal, were not royal officials and were more interested in the politics of the city than in those of the kingdom; it also represented a clear move toward both the institutionalization of local power autonomous from kings and the fragmentation of that power. In the face of the Hungarian danger, Berengar took this development one step further and localized military defense; after 900 he issued large numbers of grants to private persons, lay and ecclesiastical, of rights to build and fortify castles. His intention was carefully strategic, and his defense in depth was quite effective, but these castles in turn slowly became local centres of personalized military power, and they gained rights of private justice by the 11th century as well.

Carolingian government had always worked better when strengthened by private relationships of a political and military nature; for example, counts relied on their vassals more than on other subordinates to do their bidding, for vassals had sworn personal oaths of loyalty to them. In the castles of the 10th century, personal military bonds became the basis for effective local action. The office of count too was to become more and more the basis for private family power, particularly with the appearance in the counties during the early 10th century of newer ruling families with primarily local roots and fewer national pretensions. Cities remained important administrative centres, but they increasingly became points of reference for the family politics of the military aristocracy rather than bases for royal intervention. These processes had begun well before 888 and were not to be complete until the 11th century, but it is arguable that Berengar’s reign marked the turning point. They were crucial for the development of later urban autonomies, culminating in the city communes, but they were disastrous for kings.

Berengar gained 15 years of unopposed rule (905–921) by his cessions of rights and lands after 900. But power was slipping away. Tuscany and Spoleto were semiautonomous under their marquesses; so was the Rome of John X (pope 914–918) and of the powerful senator Marozia and her son, the princeps (prince) Alberic, who were able and effective rulers between 924 and 954. Hugh of Arles (king 926–947) found the situation irreversible. He could no longer use Carolingian-style procedures, such as new legislation or local administrative intervention, to assert his power. His most typical solution was to overthrow all obvious rivals and replace them with his own relatives, who would in theory be more loyal to him. As a result, he seemed simply violent and high-handed. But the fact is that royal power by now seemed to consist of outside intervention; kings, though still influential and rich, were outsiders to most of Italy. When Hugh faced a coup in 945, his support melted away, and he fell. When Otto I of Germany conquered the Italian kingdom, almost bloodlessly, in 962, his entirely non-Italian power base may simply have seemed to the Italians the logical conclusion of the kingship’s increasing marginality. The Italian kingdom was to survive as a coherent administrative structure at least until the 1080s, and Frederick Barbarossa even in the 1150s could seek to revive it with some success, but it was by now external to the immediate interests of most of its subjects. After Hugh, no king could establish stable power in the peninsula without a foreign power base and a foreign army.