Background: The aims were to: (i) explore the association between age and size of setting and staffing per resident; and (ii) report resident and setting characteristics, and indicators of service process and resident activity for a national random sample of staffed housing provision. Methods: Sixty settings were selected randomly from those accommodating six or fewer adults with intellectual disabilities in Wales. Nine settings subsequently withdrew. Information from the remaining 51 was obtained on resident ages, gender, adaptive behaviour, physical and sensory disabilities, social impairment and challenging behaviour, the number and working hours of staff, staff:resident ratios at different times of the day, working practices, resident receipt of attention and assistance from staff and resident activity. The sample of 51 residences was divided into four groups of 12, 13, 13 and 13 settings each, according to ascending average adaptive behaviour scale scores. Results: Age of residence was associated with larger size of residence and fewer staff hours per resident. Higher staffing per resident was associated with smaller setting size. Staffing per resident differed across adaptive behaviour groups to a limited extent. Size and age of residence, working practices, resident receipt of attention and assistance from staff did not differ across groups but resident engagement in activity did. Considerable within-group variation was found. Conclusions: The data illustrate trends in provision over time. In addition, they give a picture of recent practice and raise questions about whether staff input, on one hand, and staff training and performance, on the other hand, are well enough matched to the different needs of residents at different points of the adaptive behaviour spectrum.