The paper reevaluates the use of the ratio of nebular lines [O III] and [N II] as an abundance indicator in extragalactic H II regions, particularly for high-redshift star-forming galaxies. The authors, Max Pettini and Bernard E. J. Pagel, highlight the advantages of this method over other indices like $R_{23}$ and $N2$. They revise the compilation of data by Denicoló et al. (2002) to include only H II regions with well-determined oxygen abundances, either through the 'direct' $T_e$ method or detailed photoionization modeling. The revised data show that the relationship between $N2$ and $\log (O/H)$ is approximately linear, with a slope of 0.57, and a 95% confidence level accuracy of about 0.4 dex. For moderate to high metallicities, the $O3N2$ index, defined as $\log \{([O III] / \lambda 5007 / H \beta) / ([N II] \lambda 6583 / H \alpha)\}$, is particularly useful, showing a steep linear relationship with $\log (O/H)$ when $O3N2 < 1.9$. This index is advantageous because it has a monotonic behavior, avoids the double-valued nature of $R_{23}$, and relies on emission line ratios close in wavelength, making it insensitive to reddening and flux calibration issues. The high sensitivity of $O3N2$ to oxygen abundance also helps mitigate complications arising from rapid star formation and nitrogen deficiency in high-redshift galaxies. The authors conclude that the $O3N2$ index can provide accurate oxygen abundance measurements for star-forming galaxies at $z > 1$ with a modest investment in observing time.The paper reevaluates the use of the ratio of nebular lines [O III] and [N II] as an abundance indicator in extragalactic H II regions, particularly for high-redshift star-forming galaxies. The authors, Max Pettini and Bernard E. J. Pagel, highlight the advantages of this method over other indices like $R_{23}$ and $N2$. They revise the compilation of data by Denicoló et al. (2002) to include only H II regions with well-determined oxygen abundances, either through the 'direct' $T_e$ method or detailed photoionization modeling. The revised data show that the relationship between $N2$ and $\log (O/H)$ is approximately linear, with a slope of 0.57, and a 95% confidence level accuracy of about 0.4 dex. For moderate to high metallicities, the $O3N2$ index, defined as $\log \{([O III] / \lambda 5007 / H \beta) / ([N II] \lambda 6583 / H \alpha)\}$, is particularly useful, showing a steep linear relationship with $\log (O/H)$ when $O3N2 < 1.9$. This index is advantageous because it has a monotonic behavior, avoids the double-valued nature of $R_{23}$, and relies on emission line ratios close in wavelength, making it insensitive to reddening and flux calibration issues. The high sensitivity of $O3N2$ to oxygen abundance also helps mitigate complications arising from rapid star formation and nitrogen deficiency in high-redshift galaxies. The authors conclude that the $O3N2$ index can provide accurate oxygen abundance measurements for star-forming galaxies at $z > 1$ with a modest investment in observing time.