The relationship between vegetation and the environment is one of the fundamental questions in understanding plant species composition and community distribution along an environmental gradient. Understanding vegetation-environment correlations may help predict possible shifts in plant communities in response to climate and land use changes. There are 48 million hm² of natural grasslands on the Northern Tibetan Plateau, accounting for 59% of the total grassland area in the Tibet Autonomous Region. However, these alpine grasslands are threatened by global climate change and intense human activities in recent decades, and are now widely degraded. Quantitative classification and ordination are important methods in examining the internal classification of plant communities, characteristics of their associations, and in revealing the ecological relationships between vegetation and the environment. TWINSPAN classification and DCA,CCA ordination were used to conduct a classification and ordination of the plant communities of 29 plots in the northern Tibetan grassland. Our results indicate that: (1) Northern Tibetan alpine meadows are classified into 10 associations: Stipa purpurea+Festuca nitidula+Carex ivanovae; Stipa purpurea+Carex ivanovae; Stipa purpurea+Leontopodium nanum; Kobresia pygmaea; Kobresia pygmaea+Stipapur purea; Stipa purpurea; Stipa purpurea+Blysmus sinocompressus+Oxytropis; Stipa purpurea+Heteropappus semiprostratus+Astragalus puivinalis; Carex moorcroftii+Stipa purpurea+Rhodiola smithii; and Stipa glareosa+Stipa purpurea+Ajania khartensis; (2) The first axis of DCA basically reflected a moisture gradient, and the second axis indicated a thermal gradient; (3) Each community divided by TWINSPAN classification had its specific distribution and boundaries on the DCA ordination diagram, which indicated that DCA ordination can explain the relationships between the plots and environment; (4) CCA plots indicated that the primary environmental factors controlling the distribution of communities were rainfall (average annual rainfall) along longitude, followed by temperature (average annual temperature); CCA sorting further elaborated the relationship between plant community distribution and the environment, and indirectly supported the results of the TWINSPAN classification; (5) Quantitative analysis of 74 species by CCA indicated that the distribution pattern of species and plots in DCA was similar. The first axis of species CCA reflected water conditions of species distribution. Taken together, we conclude that the distribution patterns of plant species communities closely mirror each other. In general, the present study provides an ecological interpretation of the distribution of plant species communities along an environmental gradient on the Northern Tibetan Plateau.