Each of the jails -- Dallas County Jail, Woodlawn (minimum-security institution in the Dallas County jail system), and Harris County Detention Center -- contained housing units that varied in space per person (spatial density), number of persons living in a cell or room (social density), and number of inmates in one housing unit (external density). Each of these factors has been shown to influence inmates' reaction to housing in prisons and other environments. Data were obtained from site visits, interviews with inmates, and inmate medical records. Both social and spatial density were found to contribute to the evaluation of crowded environments. While spatial density was not a significant independent predictor in relatively spacious Woodlawn, it was a significant predictor in spatially dense Harris County Detention Center. This finding supports the hypothesis that space will become a relatively more important factor when it reaches sufficiently low levels. The effects of housing type on the illness complaint rate were not found significant. Apparently, the relatively short stays in the jails are not sufficient to build uo the substantial medical histories required for an adequate assesment of crowding effects. In prison research, the most stable findings were those for the period after the inmate had been at least 6 weeks in a particular housing unit. While blood pressure was not related to housing type in the Dallas of Woodlawn facilities, recidivists housed in the spatially dense 8-man cells at Harris had elevated diastolic blood pressures. Generally, those confined to their housing rated their housing more negatively. Yet those who were allowed to be involved in activities outside their units had elevated blood pressures. In Dallas County, units with violent inmates had more negative reactions. The results indicated that classifications should be considered in any study of jail crowding. Seven references and five notes are provided, and forms used in the study are appended.