The Newark Island Layered Intrusion (NILI) is a Middle Proterozoic (1305 ± 5 Ma; U-Pb zircon) mafic pluton on the eastern edge of the Nain Anorthosite Complex, Labrador. The NILI has been subdivided by Wiebe (1988) into a stratigraphically lower, earlier, layered series (LS) and a later hybrid series (HS). Although grossly similar in mineralogy, the REE and Sr isotope systematics of the LS and HS cumulates differ dramatically. All but one of the LS cumulates exhibit low total REE (La = 5 to 8 X chondrites) and a large positive Eu anomaly, indicative of an adcumulus to mesocumulus origin, with plagioclase and olivine as the expected major cumulus phases. Analyzed HS cumulates and other nonchilled rocks also contain principally olivine and plagioclase as cumulus phases, yet exhibit much higher abundances of the REE (La = 40 to 90 X chondrites), are LREE enriched ((Ce/Sm)n = 1.78), and show only a small positive Eu anomaly that diminishes with increasing REE abundances. No assemblage of mafic minerals alone could have generated these patterns and abundances. This suggests that HS mafic cumulates must contain a substantial amount of trapped liquid, and therefore are classified as orthocumulates. Modeling of the compatible element-incompatible element ratio Sr/Ba confirms the need for a substantial trapped liquid component in HS nonchilled mafic rocks. Sr isotopic studies indicate a restricted range for the HS cumulates (87Sr/86Srinitial = 0.7036 to 0.7040), which is indistinguishable from chilled mafic pillows (= liquids) found from within the HS. LS cumulates exhibit an extremely broad range (87Sr/86Srinitial = 0.7050 to 0.7081), which is elevated relative to any mafic rock found in the HS. The relatively elevated and variable Sr isotopic initial ratios for LS cumulates may indicate selective contamination with a more radiogenic component (crust?) in the early stages of NILI magmatism. With the intrusion of HS magmas, magmatism reached its xenith and contamination was no longer a factor. Early LS magmatism may have “armored” the conduit(s) that fed the NILI chamber and allowed later magma pulses, such as those that gave rise to the HS, to arrive at the level of the intrusion relatively uncontaminated. This scenario is analogous to what has been hypothesized for successive volcanic eruptions in the Aleutian Islands, Alaska (e.g., Myers et al., 1985; Meyers and Frost, 1994).