Patterns of coalition formation and apportionment of payoffs in two negotiable tetrad games were studied. Twenty groups were studied and male players were assigned additive power weights reflecting the value of the game. Game 1 assigned 1,2,2, and 4 to the players, and Game 2 players received 1,1,1, and 3. The winning coalition received the constant stake in the game to be divided among them according to negotiated agreement. In Game 1 any triple alliance and any alliance including the high power weight could win. Likewise, any coalition including the high weight in Game 2 could claim all of the constant payoff. However, in Game 2 the triple alliance of weak players (holders of 1's) could only tie with the high power weight. In both games, if no coalition were formed, the high power weight would claim all of the payoff. On the basis of studies of triadic groups the following hypotheses were generated: In Game 1 the high power weight was expected to gain access to pair alliances more often in later than in earlier games of a ten-game series. As this inclusion increased his payoff share was expected to become approximately equal to that of his weaker ally. In Game 2, where the strong power weight could secure half of the constant payoff whether he allied or not, it was expected that the weak triple alliance would be preferred in most groups to the winning pair which included the high weight. The findings revealed that the high power weight was included in winning pair alliances as a general thing over all time periods for both experimental games. Evidence for shifts in payoff share over time for the high power weight in Game 1 was inconclusive. Although the weak triple alliance was used above chance expectancy only in Game 2, such alliances were not preferred by most groups playing the game defined by this pattern. The difficulty of negotiating triple alliances, which require a two-step negotiation, seemed to account for the general scarcity of triple alliances. The occurrence of the weak triple alliances in Game 2 rather than in Game 1 may depend on the higher activity level of the weaker players in Game 2. Activity level was assessed by comparing frequencies of first offers in each trial for the two different games. It was proposed that differential pressures are brought to bear on the different power positions in different game-situations; however, further study of the bargaining process would be needed to establish the worth of such an hypothesis.