The effective population size (N e ) is frequently estimated using temporal changes in allele frequencies at neutral markers. Such temporal changes in allele frequencies are usually estimated from the standardized variance in allele frequencies (F c ). We simulate Wright-Fisher populations to generate expected distributions of F c and of F c (F c averaged over several loci). We explore the adjustment of these simulated F c distributions to a chi-square distribution and evaluate the resulting precision on the estimation of N e for various scenarios. Next, we outline a procedure to test for the homogeneity of the individual F c across loci and identify markers exhibiting extreme F c -values compared to the rest of the genome. Such loci are likely to be in genomic areas undergoing selection, driving F c to values greater (or smaller) than expected under drift alone. Our procedure assigns a P-value to each locus under the null hypothesis (drift is homogeneous throughout the genome) and simultaneously controls the rate of false positive among loci declared as departing significantly from the null. The procedure is illustrated using two published data sets: (i) an experimental wheat population subject to natural selection and (ii) a maize population undergoing recurrent selection.T HE effective population size (N e ), defined as the 1/2N e ) t ] (Crow and Kimura 1970). If t is not too large (t Ӷ N e ), N e can be approximated by N e Ϸ (P 0 (1 Ϫ P 0 )t)/ size of an ideal Wright-Fisher population undergoing the same rate of genetic change as the population (2V(P t )) and therefore an estimator for N e based on the standardized variance in allele frequency is (V(P t ))/ under study, is an essential parameter to predict the evolution of a population due to genetic drift in terms of (P 0 (1 Ϫ P 0 )). Nei and Tajima (1981) proposed estimating the standardized variance in allele frequency between rates of loss of genetic variation, fixation of deleterious alleles, or inbreeding (Wright 1969). However, obgeneration t x and t y for each locus l with K l alleles as taining direct estimates of N e from demographic data has often proved difficult. An alternative is to use indi-rect methods, for instance, those based on the measurement of temporal changes in allele frequencies at neuwhere p x(i,l) [respectively p y(i,l) ] represents the frequency tral markers (Krimbas and Tsakas 1971; Waples of allele i at locus l in the sample of S x individuals drawn 1989a). The foundation of these methods is that the at generation t x (respectively S y individuals at t y ). A variance of allele frequency due to drift from parents weighted mean of F c,l -values across several loci, to offspring, V(P 1 ), depends on N e as follows: V(P 1 ) ϭ P 0 (1 Ϫ P 0 )/2N e , where P 0 is the frequency in the parentalpopulation. After t generations of drift, the expected frequency of the allele is E(P t ) ϭ P 0 and the variance is then typically used to estimate N e via of the allele frequency, V(P t ) ϭ E(P t Ϫ P 0 ) 2 , can be written as a functi...