This article presents some initial findings from an ongoing research project on the way that divisional general managers of targe organizational units control new product innovations. (Control, in this instance, refers to the set of procedures, systems, and actions that general managers use to monitor, evaluate, influence, or define what subordinates are doing.) The research presented here focuses on three broad questions: 1.1. Do divisional general managers of large organizational units control their new product activities differently from their more established operations?2.2. Is a new product's innovation strategy related to the nature and degree of divisional general manager control—and, if so, in what way?3.3. Is a divisional general manager's choice of control methods related to his/her unit's new product output?The results were based on in-depth interviews with the general managers of 26 large Canadian-based divisions in 12 firms. All the firms were significant competitors in the North American market and all were actively engaged in new product activities. Firm size ranged from $210 million to $5 billion in sales. The following is a summary of the study's principal findings and conclusions: 1.1. Control varies among dimensions. The study measured the degree of control exercised by divisional general managers over new and established products on 14 control variables. It was found that none of the new products (relative to established brands) was controlled in the same fashion by the managers. Instead, new products were always managed through a variety of “loose” and “tight” controls. In so doing, it appeared that the divisional general managers were trying to balance the control and freedom required by subordinates with new products.2.2. Control varies with strategy. Both theory and empirical research generally support the notion of linkage between a unit's strategy and its organizational (eg, design, reward, placement, information, etc.) practices. The results of this study strengthen this line of thinking. The data show that both the nature and degree of divisional general manager control vary with three dimensions of product strategy (i.e., familiarity, uniqueness, and advancement).3.3. Control varies with output. The analysis highlighted the pivotal relationship between divisional general manager control and new product output. Both the nature and degree of control were found to be associated with new product output in each strategic category. Although there does not appear to be one best way to control all types of new products, some divisional general management approaches to control were more preferred than others.4.4. Loose formal/tight informal. The general managers' control patterns showed that formal control dimensions were usually managed more loosely than informal ones; that tighter informal controls were used to off-set (or “balance” ) the more relaxed formal dimensions; and that the observed reduction in formal control should not be interpreted to mean either the absence of bureaucracy or the absence of formal control. Indeed, some formal bureaucratic control was always found in the “high output” strategic categories. Thus, rather than being considered or labeled as typically “bad”, bureaucracy may in fact be “beautiful”—provided, of course, that it is appropriately used.The article concludes by arguing that a divisional general manager's approach to controlling new products seems to make a difference in terms of performance. As such, the control approach chosen should not be made haphazardly or with abandon.