This study investigated the effects of task type (production, discussion, and problem solving), group sex composition (three males, two males and one female, one male and two females, and three females), and ordinal position of task presentation (first, second, third, and fourth) on characteristics of written group products. Four groups of each sex composition completed four tasks of a single task type, yielding 192 group products. Products were rated by four judges on six general product dimensions: action orientation, length, originality, optimism, quality of presentation, and issue involvement. In addition, products were judged on two task-dependent dimensions, adequacy and creativity. Task type and specific task effects were most important as determiners of group product characteristics. Sex composition, its interactions, and specific group effects were of only moderate importance. Ordinal position and its interactions were of minimal importance. In general, task type effects replicated those reported by Hackman 1968. Specific task and specific group effects found in this study were supported by a re-analysis of the Hackman 1968 data. Implications of the strong task effects for future research were discussed. © 1969.