Both heightened and attenuated stress reactivity from prior stress also have been documented in other mammals (Koolhaas et al., 2006). Milder stressors tend to attenuate later stress responses, whereas trauma tends to enhance them. However, the same stressor can affect various stress response systems differently (e.g. sympathetic versus hypothalamic-pituitary-adrenal activity, Schommer et al., 2003). Futhermore, situational and dispositional variables moderate the impact of prior stress on subsequent vulnerability (e.g., Gunnar and Vasquez, 2006; Moore et al., 2006). A useful paradigm for studying sequential stress effects in laboratory rats centers on modulation of acoustic startle amplitude. Startle is a defensive reflex, and startle testing after earlier stress constitutes a mild stressor reexposure (Commissaris et al., 2004; Cranney, 1988). Prior experimental stress, such as inescapable shock, usually sensitizes startle. Sensitization has been observed seconds (Pilz et al., 1999), minutes (Davis, 1989), and days (Servatius et al., 1995) after shock. As stressor severity increases, sensitization emerges later and lasts longer. Sensitization emerges a few minutes after ten 0.5-s, 0.6-mA footshocks and is gone within 40 min; at 1.0 or 1.4 mA, sensitization emerges in 20-30 min and is robust at 40 min (Davis, 1989). Sensitization is detectable a week after one session of forty 3-s, 2-mA tailshocks, but not until 10 days after three sessions (Servatius et al., 1995; see also Matuszewich et al., 2007). The degree to which shock sensitizes startle also varies among rats. Sprague-Dawley rats sensitize more than Wistars (Pilz et al., 1999; also see, Faraday, 2002). Using startle nonhabituation after shock as a PTSD model, Garrick et al. (2001) observed nonhabituation only in rats with low baseline startle amplitudes. Milde et al. (2003) also found sensitization after shock in a subset of rats, those with low plasma corticosterone (CORT). Compared to sensitization, stress-induced startle attenuation seems rarer and more circumscribed. Beck and Servatius (2005, 2006) observed attenuation 2 h after shock but not after restraint, and only in intact females, concluding that attenuation depends on nociception and ovarian hormones. However, Conti and Printz (2003) observed attenuated startle 24 h after restraint in males of some rat strains. Stress-attenuated startle appears to be a potentially important phenomenon in need of further study. Accounting for individual differences in stress-induced changes in startle requires approaches in which dispositional variables are well characterized. Selective breeding is such an approach: Lines are developed on a clearly operationalized phenotype, after which phenotypic correlates and functional relationships are examined. For example, Roman low (RLA) and high (RHA) avoidance rats are