OBJECTIVE Although respiratory care is a relatively new profession, its practitioners are deeply involved in providing patient care in the critical care. In preparation for writing this article, we sought to explore the respiratory therapy manpower needs and activities designed to fulfill those needs in critical care practice. METHODS We began by delineating the historical development of respiratory care as a profession, the development of its education, and the professional credentialing system. We then conducted several literature reviews with few articles generated. We requested and received data from the American Association for Respiratory Care (AARC), The National Board for Respiratory Care (NBRC), and the Committee on Accreditation of Respiratory Care education (CoARC) relative to their membership, number of credentialed individuals, and educational program student and graduate data for 2000 through 2004. We then conducted two electronic surveys. Survey 1 was a six-item survey that examined the use of mandatory overtime in respiratory care departments. We used a convenience sample of 30 hospitals stratified by size (<or=200 beds, 201-499 beds, >or=500 beds). Survey 2 was a five-item instrument distributed by blast E-mail to the Society of Critical Care Medicine's Respiratory Care Section members and members of the RC_World list serve. This survey elicited 51 usable and non-duplicative responses from geographically and size-varied institutions. We analyzed these data in several ways from distribution analysis to one-way analysis of variance procedure and appropriate post hoc analysis techniques. Where appropriate, a matched-pairs analysis was performed and these were compared across the variables intensive care unit (ICU) beds per actual number of respiratory care practitioners (RCPs) and ICU beds per preferred number of RCPs. RESULTS The data gathered from the professional organizations indicated a relatively stable attrition rate (35.2%+/-1.7-3.1%), even in the face of varying enrollments (6,231 in 2004 vs. 4,589 in 2002). In survey 1, we looked at the institution of mandatory overtime policies and their use in 30 size-stratified hospitals. Mandatory overtime was selected as a survey topic under the supposition that manpower shortages might lead to the development of such procedures and also to their utilization. Fourteen of the 30 hospitals responding indicated that they had a policy addressing mandatory over time. Of the 14 hospitals with policies, only ten had disciplinary actions specific to refusing the overtime. Seven of the 30 hospitals indicated that they used mandatory overtime monthly of more frequently. Survey 2 data revealed that there was a wide variation in bed size, number of ICUs, and number of RCP staff assigned to the ICU. Serendipitously, our 51 responding centers were distributed among small (16), medium (19), and large (16) hospitals in a manner that appeared to reflect the national distribution pattern. We were able to use these data to develop a closeness of fit diagram ICU beds to preferred numbers of RCPs (DF=48; p<.0001; RSq=0.77; RMSE=4.114). The number of beds per preferred number of RCPs was 9.445 to 1.0 while the actual bed to RCP ratio was 10.75 to 1. CONCLUSIONS This article provides a short history of the development of respiratory care and its historical relationship with critical care. We have, perhaps for the first time, provided a unified data set of key demographic information from the three professional bodies guiding the development of the respiratory therapy profession. This data set provides time-linked data on admissions and graduations from the CoARC, membership numbers for the AARC, and the numbers of active credentialed RCP from the NBRC. By two focused surveys, we were able to show that while mandatory overtime is a common practice in respiratory care departments, it was not overwhelming utilized. We also learned that in most hospitals, regardless of bed size, there is a perceived need for 1.3 RCPs more than the actual staff and that it appears that the critical staffing level between actual to preferred RCP to beds is between 9 and 11 beds.