The Community Integration Program Questionnaire was developed to measure various quantifiable characteristics of community integration programs for people with brain injury. There are three versions: one for outpatient facility-based, one for residential and one for home programs. In this study questionnaires were administered to directors and associate directors of seven programs. A research assistant then went on-site to collect corresponding data using patient records, planning books and schedules, annual and quarterly reports, employee lists, staff curricula vitae, census data, insurance payment data, team meeting notes, incident reports, and staff logs. Matching criteria between questionnaire and on-site data were established before either was collected. The survey data matched data acquired on-site in 42.2, 56.4, and 56.8% of questions in the three versions, respectively. Interviewers and a second listener recorded the same information 91.5% of the time across seven program interviews. For two programs, comparisons between a Program Director and Program Coordinator yielded matches on 42.8 and 40.6% of questions, respectively. In conclusion, the Community Integration Program Questionnaire does not have acceptable construct validity and inter-rater reliability. Researchers requiring information about quantifiable characteristics of community integration programs should go on-site to collect the data.