Several experiments were conducted to test whether, as suggested by Welch et al. in this journal, mere group living (social stimulation) can account for the significant differences in measures of brain anatomy and brain chemistry that develop between rodents housed in groups in enriched environments and rodents housed singly in restricted environments; the alternative hypothesis was that features of the inanimate environment can significantly affect brain measures of animals living in a social group. Groups of 12 male rats were assigned for 30 days to several types of environment: (a) large cage without stimulus objects, (b) large cage containing varied stimulus objects, (c) large cage containing a maze whose pattern of barriers was changed daily, and (d) a seminatural outdoor environment; in each experiment, littermates of rats in the social conditions were housed in isolation in small colony cages. At the end of the 30-day period, measures were taken of weights of brain regions, RNA and DNA contents of regions of cerebral cortex, and acetylcholinesterase activities of brain regions. Although the number of rats housed together was constant for conditions a--d and cage size was constant for conditions a--c, the magnitudes of the cerebral measures varied significantly as a function of the inanimate stimulus conditions. The differences from isola;ion-housed littermates was greatest in condition d and smallest in condition a. Thus, social grouping alone is inadequate to explain the cerebral effects of enriched environments and the inanimate stimulus conditions must be taken into account.