Researchers often use more than one genetic method to estimate contemporary effective population size (Ne), but few formally combine multiple estimates despite the potential benefits for increasing precision. Maximizing these benefits requires an optimal, inverse-variance weighting scheme. A precondition for combining estimates is that they must be estimating the same parameter, which can be appropriate either for estimates using the same method applied to different time periods, or for estimates using different methods applied to the same time period. Previous approaches focused on var(Ne^) for weighting, but that is problematical because Ne^ is highly skewed and can be infinitely large. A new approach is described here using weights that are inversely proportional to var(1/Ne^). 1/Ne^ is the drift signal that Ne-estimation methods respond to, and its distribution is close to normal even when Ne^ assumes extreme values. Benefits are maximized under three general conditions: the estimators have approximately equal variances; they are uncorrelated or have weak positive correlations; individual estimates have low precision (i.e., if data are limited and/or true Ne is large). Analytical and numerical results demonstrate that: (1) existing theory allows robust estimates of var(1/Ne^) for the temporal and LD methods, which provide independent information about Ne -- both of which facilitate optimally combining those methods; (2) estimates for the LD and sibship methods are essentially uncorrelated when data are limited but can be strongly positively correlated in genomics-scale datasets. General theory predicting var(1/Ne^) for the sibship method is lacking, but values for specific scenarios have been published.