In a model plant species Arabidopsis, one of key determinants for flowering time is a MADS-box floral repressor, FLOWERING LOCUS C (FLC). FLC is silenced after a sufficient period of winter cold has been perceived (known as vernalization response). The lack of FLC expression allows plants to achieve the competence to flower in spring through the activation of floral integrator genes. Previous studies revealed that repression of FLC by vernalization is achieved in part by an evolutionarily conserved chromatin modifying complex, Polycomb repressive complex 2 (PRC2). In Arabidopsis, PRC2 (which contains CLF, an E(z) homolog, a H3K27 methyltransferase) is recruited to FLC chromatin upon vernalizing cold and mediates methylations at Histone H3 Lys 27 (H3K27me3), a repressive histone modification mark. Recent reports identified a plethora of long and/or short noncoding RNAs (ncRNAs), which contribute to the recruitment of PRC2 to its target chromatins. In vernlaization, a long noncoding RNAs (lncRNA), named as COLDAIR, was shown to mediate FLC silencing by vernalization. COLDAIR lncRNA binds directly to CLF, a PRC2 component, and is necessary for increased enrichment of PRC2 at FLC chromatin by vernalization. In addition, we have identified additional noncoding RNAs that are involved in various developmental processes in plants. Using the FLC regulation as a model, I will discuss our current understandings on epigenetic FLC silencing by protein and noncoding RNA components. As an alternative model system, we also have studied maize endosperm development to dissect epigenetic mechanism governing its developmental processes. I will briefly discuss our progress on the understanding of maize endosperm development as well.