This study shows that there was a shift in political and military power in the Safavid state, for which the basis was laid in the mid-16th century. The shift itself was formalized after 1590 through the reforms initiated by `Abbas I. The changes took three different forms. First, the monopoly on military power of the Qezelbash tribes was broken by severing the direct ties between the Qezelbash leaders and their clansmen. This was achieved by dislodging the Qezelbash leaders from their traditional fiefdoms and appointing them as governors in other areas. Furthermore, captured or bought slaves (gholams) were trained to serve both as a military countervailing force as well as to serve the shah in the country’s administration. The power shift did not mean that the Qezelbash elite were ousted from power. However, it meant that their monolithic hold on power was broken as was their oligopoly of military might. Henceforth, they had to share power with the gholams. The qurchi-bashi, the military leader of the shah’s household troops, became the leader of all Qezelbash forces. Although he became the most important military leader, his powers were checked by those of the qollar-aghasi, the commander of the gholams. The latter sometimes also held simultaneously other powerful central government functions, such as that of tofangchi-aghasi and divan-begi. The former function had been traditionally held by Tajiks and the latter by Qezelbash. However, the gholams did not have an exclusive hold on any of these three functions, for throughout the 17th century and the beginning of the 18th century, these functions were also held by Qezelbash emirs. There were other contenders for power as well. As part of the power shift, the Tajiks, who traditionally had held mainly the highest administrative functions, now lost exclusive hold on those functions, where traditionally they had held a monopoly. This was so not only for functions such as that of nazer-e boyutat or the steward of the royal palace, but also for lower-ranking administrative functions. The biggest loss that the Tajiks suffered was their exclusive claim on the function of grand vizier, which as of 1669 was held almost uninterruptedly by Qezelbash officials till the end of the regime. The gholams and Tajiks were not the only social group that vied for a piece of the power pie. The harem inmates increasingly also wanted to have their say. Royal women had always played a role behind the scenes in Safavid politics in the 16th century, while, on occasion, they even tried to hold outright power directly (sister of Tahmasp II; wife of Khodabandeh). Generally, however, the royal women could not play a direct role in the administration of the country, but their fellow harem inmates, the eunuchs, could. They were the only ones, apart from the shah, who had direct contact with the harem, and they could hold office. Therefore, we see that eunuchs, apart from trying to influence state policy from behind the harem walls, also participated directly in the administration of the country. For example, as of the 1680s, eunuchs held functions such as that of nazer-e boyutat, which had been a Tajik prerogative until the 1650s, and of jabbehdar-bashi, which always had been held by the Qezelbash. It is clear that the new configuration of power-sharing between the contending members of the elite was to the advantage of the shah because the various opposing forces held each other in check. It meant, however, that the political decision-making process became more difficult. Finally, it led to paralysis when Shah Soltan Hoseyn came to power. He was so weak that he agreed with every last person who talked to him. Rather than allowing one’s opponent to have his way, even when it would be in everybody’s interest, the opposing elite members preferred that nothing happened at all to avoid one's rival gaining political or other advantage. Second, there was a shift in power between those officials who managed the mamalek or state lands and those who managed the khasseh or household lands. Because of the increase in the proportion of khasseh lands, the officials in charge of those properties saw their role and importance increase. This was but a logical outcome of the reduction of power of the Qezelbash, whose economic power base had been the governing of the mamalek lands. The change in the number and size of mamalek lands reduced their ability to finance troops and, thus, diminished their military might. Because of the growth of the khasseh lands in the 16th century, which were governed mainly by royal viziers, and because the land’s revenues were remitted directly to the royal court, the powers of the managers of the khasseh lands increased accordingly. Therefore, we see that invariably those officials, whose importance had been subordinate to that of their mamalek colleagues during the 16th century, reversed that role during the 17th century. For example, the monshi al-mamalek had to yield to the majles-nevis and the sadr-e mamalek to the sadr-e khasseh. Only the mostowfi-ye khasseh still remained subordinate to the mostoufi-ye mamalek, although he played a much more important role than before. This change in circumstances was reflected in administrative procedures. Whereas in the 16th century the neshan and parvaneh (both mamalek documents) had been predominant, in the next century it was the raqam, a khasseh document. Also, whereas the former required the counterseal of the Keepers of the Seal (a function held by Qezelbash), the latter was sealed by the shah. Third, apart from changes, there also was considerable continuity in that both the function of the shah and the grand vizier remained constant. The position of the shah, who enjoyed enormous legitimacy, was never seriously challenged. The shah’s means to exercise power, through the growth of his hold on the means of production in the country, had increased considerably, however. Therefore, the elite tried to undermine the position of the shah’s main agent, the grand vizier. For it was the grand vizier who continued to hold the purse strings and the keys to political influence from the beginning to the end of the Safavid reign. Whereas in the 16th century the Qezelbash emirs tried to impose themselves on the grand vizier, in the 17th and 18th centuries, they were joined by the gholams, Tajiks, and harem inmates. This shows that there was significant continuity in the authority of the grand vizier, but that the game had remained the same and that only the composition of the players had changed. The fact also that a gholam, between 1655-61, and Qezelbash officials, as of 1669, regularly held the office of grand vizier not only demonstrates their absorption into the Tajik dominated bureaucratic system, but also shows that the players rather than trying to beat the system had decided to join it. It was, of course, still possible to weaken the position of the grand vizier, but it was only a temporary gain, for the shah, if he did not dismiss the grand vizier, when push came to shove always backed his own man.