The presence of allocatable fixed inputs may cause truly joint technologies to appear nonjoint in the short run as well as truly nonjoint technologies to appear joint. This paper demonstrates theoretically why this can happen and then documents that it actually occurs in a significant way in aggregate U.S. agricultural production. A simple testing procedure is used that requires no data on inputs allocations. The important finding is that failure to reject true (apparent) nonjointness does not justify modeling short-run (long-run) supply independent of alternative output prices.