The Texarkana Federal Correctional Institution (TFCI) and Dallas County Jail (DCJ) had not only the same administrative milieu, but also architectural features which provided a wide range of cell and dormitory sizes and population levels which would give substantial variation in the two basic components of crowding, spatial density (the amount of space per inmate) and social density (the number of inmates within a housing unit). To put the crowding figures in perspective, it should be realized that a 1,000 square foot house would have to hold approximately 50 residents to approximate the spatial and social density found at the DCJ. Initial study results indicated that tolerance for crowding in high social density conditions declined as a function of length of time in such conditions, whereas no relationship was found for single cell housing. Subsequent studies have investigated the possibility that high social density of the kind encountered in prisons and jails would generate indications of psychological stress such as apparent in illness complaints and increased palmar sweat. In both TFCI and DCJ, illness complaint rates were greater in more crowded conditions. Palmar sweat prints obtained from 46 TFCI inmate volunteers showed a significant positive correlation of palmar sweating with social density. The relationship between spatial density and palmar sweating was negligible. Although the study has shown that dorms of 26 or more men significantly increase stress levels over units of 1 or 2 men, it is possible that small dorms of 6 to 10 men may not be very stressful. Further studies will be required to determine whether crowding in prisons also leads to the social withdrawal, lack of perceived control, feeling of social overload, and disturbance of cognitive functioning in other settings. The findings of sex differences in response to crowding are also worthy of further study in a prison setting. The effects of visual contact with inmates in other cells also needs to be investigated. Over 30 references are appended.