1466American Journal of Botany 95(11): 14661474. 2008.The patterns of genetic variation within and among popu-lations are of interest to diverse fields in plant biology in- cluding population genetics, systematics, and conservation. For the past four decades, following the demonstration of the utility of enzyme electrophoresis ( Harris, 1966 ; Hubby and Lewontin, 1966 ; Lewontin and Hubby, 1966 ), there has been an ever-increasing use of various types of molecular markers to assess genetic variation. The basic rationale for molecular markers replacing earlier approaches such as quantitative characters as a means of assessing genetic variation is the more direct equation between genotype and phenotype ob- tained with molecular methods ( Lewontin and Hubby, 1966 ; Schulman, 2007 ). However, as the science has progressed, considerations of improved efficiency and sensitivity have promoted the development of new molecular markers, many of which present significant analytical challenges to accu- rately assessing genetic variation ( Sunnucks, 2000 ). Al- though ongoing developments in statistical analyses offer the potential to overcome these challenges, there is still a gen- eral inability to knowledgeably compare analogous estimates derived from different marker classes and/or statistical meth- ods ( Bonin et al., 2007 ). For more than four decades, allozyme markers have been an invaluable tool for studies of evolutionary genetics, providing plant biologists with a straightforward, low cost means of esti-mating levels of intraspecific genetic variation ( Cruzan, 1998 ). Allozymes produce codominant data, which permit direct ob- servation of allele frequencies at allozyme loci and can be used rather simply to calculate various gene statistics ( Hubby and Lewontin, 1966 ; Lewontin and Hubby, 1966 ; Hamrick, 1989 ; Weeden and Wendel, 1989 ). In addition, because of the highly conserved nature of allozyme loci in flowering plants ( Gottlieb, 1982 ), homologous loci can be compared between closely re- lated species. Practical advantages of allozymes include the relative procedural simplicity and low cost of the method ( Clegg, 1989 ). Because of the large database that has accrued for allozymes, their estimated patterns of variation can be com- pared among plants with different ecological and life history traits (e.g., Hamrick and Godt, 1989 ). One of the major criti- cisms of allozyme data concerns the level of genome sampling; allozyme variation can only be determined for protein-coding genes (many of the assays are for enzymes of glycolysis and the citric acid cycle) of which, in plants, there is a rather small (40) potential pool of useful candidates ( Clegg, 1989 ; Wendel and Weeden, 1989 ). This number is further reduced for within- species studies where often only about 50% of the loci are poly- morphic ( Hamrick and Godt, 1989 ). In addition, variation at each of these loci may only be detected if it affects the electro- phoretic mobility of the enzyme with the standard conditions employed ( Lewontin and Hubby, 1966 ; Clegg, 1989 ). As much as 20% of the base substitutions may go undetected ( Coates and Byrne, 2005 ). Allozyme variation is often absent in groups of recently radiated taxa for which allozymes frequently provide limited and/or imprecise estimates of population genetic struc- ture (e.g., Schwartz, 1985 ; Crawford et al., 1987 ). 1 Manuscript received 9 March 2008; revision accepted 25 August 2008. The authors thank K. Holsinger, M. Holder, and J. Kelly for discussions of Bayesian methodology and its application to the current study. This research was supported by the Department of Ecology and Evolutionary Biology and the Natural History Museum and 3 Jard n de Aclimataci n de la Orotava, Puerto de la Cruz, Tenerife, Canary Islands, Spain Accurate determination of patterns of genetic variation provides a powerful inferential tool for studies of evolution and conser- vation. For more than 30 years, enzyme electrophoresis was the preferred method for elucidating these patterns. As a result, evo- lutionary geneticists have acquired considerable understanding of the relationship between patterns of allozyme variation and aspects of evolutionary process. Myriad molecular markers and statistical analyses have since emerged, enabling improved esti- mates of patterns of genetic diversity. With these advances, there is a need to evaluate results obtained with different markers and analytical methods. We present a comparative study of gene statistic estimates ( F ST , G ST , F IS , H S , and H T ) calculated from an in- tersimple sequence repeat (ISSR) and an allozyme data set derived from the same populations using both standard and Bayesian statistical approaches. Significant differences were found between estimates, owing to the effects of marker and analysis type. Most notably, F STestimates for codominant data differ between Bayesia