||Background . Nursing leadership is identified as a critical aspect of assuring safe, quality care for patients in hospitals (IOM, 2004). The superiority of magnet hospital environments with regard to quality and safety outcomes has been documented (Aiken, Smith & Lake, 1994; Aiken, Sloane & Klocinski, 1997; Aiken, Sloane, Lake, Sochalski & Weber, 1999; Clarke, Rockett, Sloane & Aiken, 2002a). While it has been suggested that the nursing leadership in magnet hospitals is consistent with transformational leadership (TFL) (McDaniel & Wolf, 1992; Upenieks, 2003a), TFL in magnet v. non-magnet hospitals had not been statistically compared. Purposes . The purposes of this study were to compare staff nurse perceptions of the TFL behaviors of nurse executives (NE) and nurse managers (NM) in magnet hospitals with those in non-magnet hospitals. The study also sought to determine if there is a "cascading effect" (Bass, Waldman, Avolio & Bebb, 1987) between the TFL behaviors of NE and NM in both magnet and non-magnet hospitals as perceived by staff nurses. Methods . The study utilized a quantitative comparative and correlational design and employed secondary analysis of an existing data set originally collected for a study conducted by a New Jersey hospital health network. Content validity testing was used to evaluate the Senior Leadership and Supervisor subscales of the 2003 Press Ganey Employee Perspectives(TM) Survey for the TFL behaviors identified by Kouzes and Posner (2002) and found items to be content valid for three of the five TFL behaviors in the NE and four of the five TFL behaviors in the NM. Results . Findings demonstrated stronger staff nurse perception of NE (t = 16.301; p < .005) and NM (t = 17.477; p < .0005) leadership in the non-magnet sample (n = 470) than in the magnet sample (n = 235). Ancillary analyses designed to determine if the operational definitions used for magnet and non-magnet hospitals contributed to these findings were conducted. This re-categorization of hospitals did not affect the results. Findings also indicated moderate but statistically significant relationships between staff nurses' perceptions of TFL behaviors of the nurse executive and nurse manager behaviors in the magnet ( r = .373; p < .05) and non-magnet (r = .442; p = .01) hospital samples. Discussion . Staff nurses' perceptions of both NE and NM TFL were stronger in the non-magnet hospital sample than in the magnet hospital sample. While demonstration of a quantitative difference in staff nurse perceptions of NE and NM TFL between magnet and non-magnet hospital samples was expected, the direction of the differences was surprising. Limitations that may have contributed to the unanticipated results included: the instrument was not representative of the full range of TFL behaviors in either the NE or NM; sampling limitations of the original data set; and confounding variables not addressed in the original data set and thus unavailable for secondary data analysis. The findings also provide empirical support for the conclusion of a "cascading effect" relationship between staff nurses' perceptions of NE and NM TFL in magnet and non-magnet hospital settings. The results of the study may provide insights to nursing educators for the development of curricula and for nursing administrators as they cultivate strategies for the identification and development of TFL behaviors in nurses aimed at supporting the provision of safe, quality care to patients in hospitals.