The paper introduces a new method for analyzing spatial autocorrelation in genetic data, focusing on multiallelic codominant loci and multilocus arrays. The authors develop a general multivariate method based on genetic distance measures, illustrate its application with an example from the orchid *Caladenia tentaculata*, and provide nonparametric permutational testing procedures. The method treats the genetic dataset as a whole, reducing stochastic noise and strengthening the spatial signal. Key findings include: 1. **Multivariate vs. Single-Allele Treatment**: The multivariate treatment outperforms single-allele analysis, especially for rare alleles. 2. **Multilocus vs. Single-Locus Analysis**: Multilocus analysis provides clearer results than separate single-locus analyses, with a smoother curve and better detection of spatial structure. 3. **Weighting Alleles**: Differential weighting of alleles does not significantly improve resolution, though it may help in certain cases. 4. **Application to Different Genetic Markers**: The method is applicable to various types of genetic markers, including dominant/recessive RAPDs and haplotypic markers. The authors conclude that their method offers a robust and flexible approach to spatial autocorrelation analysis in genetic studies, particularly for multiallelic and multilocus data.The paper introduces a new method for analyzing spatial autocorrelation in genetic data, focusing on multiallelic codominant loci and multilocus arrays. The authors develop a general multivariate method based on genetic distance measures, illustrate its application with an example from the orchid *Caladenia tentaculata*, and provide nonparametric permutational testing procedures. The method treats the genetic dataset as a whole, reducing stochastic noise and strengthening the spatial signal. Key findings include: 1. **Multivariate vs. Single-Allele Treatment**: The multivariate treatment outperforms single-allele analysis, especially for rare alleles. 2. **Multilocus vs. Single-Locus Analysis**: Multilocus analysis provides clearer results than separate single-locus analyses, with a smoother curve and better detection of spatial structure. 3. **Weighting Alleles**: Differential weighting of alleles does not significantly improve resolution, though it may help in certain cases. 4. **Application to Different Genetic Markers**: The method is applicable to various types of genetic markers, including dominant/recessive RAPDs and haplotypic markers. The authors conclude that their method offers a robust and flexible approach to spatial autocorrelation analysis in genetic studies, particularly for multiallelic and multilocus data.