Fine-scale spatial genetic structure, which indicates nonrandom spatial distribution of genotypes or genetic diversity, has important consequences for population biology. The study of fine-scale spatial genetic structure can provide an understanding of the key processes and mechanisms involved in the maintenance of plant populations. Glechoma longituba is a perennial herbaceous clonal plant species that belongs to the Labiatae family. Glechoma longituba is a herb of medicinal importance that is widely distributed in China, and its phenotypic characteristics vary among different habitats. The genetic diversity, clonal diversity, and fine-scale spatial genetic structure of Glechoma longituba plants collected from three different patches (Shuiqubian, Pinggecun, and Zhulinxia) with different habitats were analyzed using inter-simple sequence repeat (ISSR) molecular markers. In addition, the correlation with habitat heterogeneity, propagule propagation, and human disturbance were also examined in the study. The results indicated the following: 1) Genetic diversity of Glechoma longituba at the species level was relatively low (percentage of polymorphic loci, P=31.15%; Shannon informative index, I=0.1601; Nei's index, h=0.1096). Genetic diversity of Glechoma longituba was highest in Shuiqubian patch (P=21.31%, I=0.0965, h=0.0627), followed by Pinggecun patch (P=8.20%, I=0.0354, h=0.0226), and Zhulinxia patch (P=3.28%, I=0.0120, h=0.0073). 2) Clonal diversity of Glechoma longituba at the species level was relatively high (number of genets, G=73; ratio of genets to ramets, G/N=0.2332; Simpson's diversity index, D=0.8843; genotypic evenness, E=0.8192). The clonal diversity of Glechoma longituba was highest in Shuiqubian patch (G=60, G/N=0.5660, D=0.9693, E=0.8747), followed by Pinggecun patch (G=10, G/N=0.1087, D=0.8430, E=0.9075), and Zhulinxia patch (G=3, G/N=0.0260, D=0.2642, E=0.3599). 3) Genetic differentiation coefficient (G_st) was 0.7129, which indicated that most of the genetic variation existed among patches, whereas little genetic variation existed within patches. The estimated gene flow was as low as 0.2004. 4) Spatial autocorrelation analysis showed that the autocorrelation coefficient of Glechoma longituba in Zhulinxia patch was significantly positive at a distance of 100 cm with an X-intercept of 205.994 cm but significantly negative at a distance of 350 cm. The autocorrelation coefficient in Pinggecun patch was significantly positive at a distance of 200 cm with an X-intercept of 235.388 cm but significantly negative at a distance of 450 cm. The autocorrelation coefficient in Shuiqubian patch was significantly positive at a distance of 150 cm with an X-intercept of 240.336 cm but significantly negative at a distance of 350 cm. Analysis with SPAGeDi 1.2 software showed that the strength of spatial genetic structure in Pinggecun patch was greater than those in Shuiqubian and Zhulinxia patches. The Sp ratio (used to compare the extent of spatial genetic structure among populations) for Pinggecun, Shuiqubian, and Zhulinxia patches was 0.0944, 0.0558, and 0.0556, respectively. The genetic diversity, clonal diversity, and fine-scale spatial genetic structure of Glechoma longituba are affected by propagule dispersal characteristics, human disturbance, and trade-off between investment in sexual reproduction and clonal propagation and might be a consequence of adaptation to habitat heterogeneity.