The ``Baldwin Effect'' is an empirical trend for decreasing line equivalent widths with increasing luminosity (Baldwin 1977, Kinney et al. 1990, Osmer & Shields this volume). Calculations by HF93a showed that most of the line equivalent widths will decline naturally with increasing metallicities (because of the declining gas temperatures, see also HF99). If Z is correlated with the luminosities of QSOs, as suggested by the emission line data (Section 4.2), then metallicity difference would at least contribute to the Baldwin Effect. An important test of the influence of metallicity will come from the nitrogen lines, because they should run counter to the trend and get relatively stronger with Z (and L) due to the selective N enhancement (Section 2.4 and Section 3). Recent observations support this prediction, showing that while CIV and other lines decline with L - the usual Baldwin effect - NV does not (Osmer et al. 1994; Laor et al. 1995, Korista et al. 1998). See Korista et al. (this volume) for further discussion.