INTRODUCTION In terms of annual production, rice (Oryza sativa L.) is the second most important staple crop worldwide, only behind wheat. Rice is a member of the genus Oryza of the family Poaceae. Twenty-two wild species as well as two cultivated species are found in the genus Oryza, namely Oryza sativa and Oryza glaberrima. There are 24 chromosomes in the AA genome of Oryza sativa, which is a diploid species. In India, rice crops account for approximately 22 percent of the total cropped area under cereals, and about 31 percent of the total area under food grains (Singhal, 2003). Rice production in Asia is more than any other cereal, and Asia is considered the "Rice Bowl" of the world (Lu and Chang 1980). The area under rice cultivation in India is 42.56 million hectares (ha), which is equivalent to 95.33 million tonnes of rice produced at a productivity of 2240 kilograms per hectare. Rice is cultivated throughout India. Breeding programs' primary objective is to develop rice cultivars with the maximum yield potential (Jing and Jianchang 2011). It is crucial to determine how different traits contribute to yield. The knowledge of genetic variability of yield contributing characters, their inter relationships, and how those relationships influence yield is essential for an effective breeding program. Heritability has been used in selection as it indicates how transmissible a trait is in future generations (Sabesan et al., 2009). Breeding programs that aim to develop high yielding cultivars must consider genetic variation for traits under selection and high heritability and genetic advancement (Ulloa, 2006). Genetic progress is not reflected by estimating heritability alone (Chaghakaboodi et al., 2012). In addition, genetic advance is of considerable significance since it indicates the degree of improvement expected from one selection (Hamdi et al., 2003). An integrated function of many component traits, grain yield is a complex characteristic. As a result, selecting for yield may not be as satisfactory as considering other traits that contribute to yield (Akinwale et al., 2011). There has been very little work done on rice grown on upland areas, which led to this study being undertaken to examine the variability of the genetic makeup of upland rice and the traits that affect yield. Therefore, the objective of this study was to estimate genetic variability, heritability, and genetic advance and to determine the correlations among yield characteristics and yield attributing characteristics through use of phenotypic correlation coefficients. The results of this study could serve as a basis for developing desirable rice genotypes for upland environments in future breeding programs. MATERIALS AND METHODS Thirty-eight diverse genotypes viz. PHY-201, PHY-202, PHY-203, PHY-204, PHY-205, PHY-206, PHY-207, PHY-208, PHY-209, PHY-210, PHY-211, PHY-212, PHY-213, PHY-214, PHY-215, PHY-216, PHY-217, PHY-218, PHY-219, PHY-220, PHY-221, PHY-222, PHY-223, PHY-224, PHY-225, PHY-226, PHY-227, PHY-228, HUR-3022, PHY-230, PHY-231, PHY-232, PHY-233, PHY-234, PHY-235, PHY-236, PHY-237, PHY-238, PHY-239, PHY-240 were collected from All India Coordinated Research Project on Rice and were used as experimental materials. At the Agricultural Research Farm of Institute of Agricultural Sciences, Banaras Hindu University, Varanasi, Uttar Pradesh, the experiment was conducted in randomized block design (RBD) with three replications. Varanasi is located at 25.18° N latitude and 83.03° East longitude, at an altitude of 123.3 meters above mean sea level. During the experiment, the soil was classified as an Indo-Gangetic plain type soil and was characterized as fertile, alluvial loam. A total of 1100 mm of rain falls annually at this place, 88 percent of which is received between July and September. Twenty one days old seedlings were transplanted into small plots 20 cm apart between rows and 15 cm apart within rows. A healthy crop was raised by following standard cultural practices. Per hectare, 120 kg of N, 80 kg of P2O5, and 80 kg of K2O were applied as fertilizers. IRRI standard evaluation system was followed for the collection of morphological data at appropriate growth stages. Other intercultural operations were carried out whenever required. Data Collection: Observations were recorded on ten randomly selected plants from each entry and replication on days to 50% flowering (D50F), days to maturity, plant height (cm), days to panicle initiation (DPI), Days to first flowering (DFF), panicle length (cm) (PL), total number of tillers (TNT), number of effective tillers plant–1 (NET), number of filled grains panicle–1 (Gr/P), Seed index (SI), Kernel length (KL), Kernel breadth (KB), Kernel L:B ratio, Grain weight per plant (GW/Pt). Statistical Analysis: Based on randomly selected plants from all three replications, the experimental data were calculated by their mean value. According to the procedures outlined by Johnson et al. (1955); Burton, (1952); Allard (1975), the genotypic and phenotypic coefficients of variation were determined statistically. Based on Johanson et al. (1955); Hanson (1961), heritability and genetic advance were estimated. Correlation coefficient was calculated according to Al-Jibour et al. (1958). RESULTS AND DISCUSSION According to the analysis of variance, there were significant differences between genotypes for all the thirty eight genotypes of characters investigated. As a result, there are many potential genotypes from the current germplasm set that can be selected for yield improvement. Per se performance: PHY-223 and PHY-227 were the genotypes that achieved the greatest mean grain yield per plant of all genotypes analyzed. These genotypes may be utilized in a hybridization programme to achieve higher grain yield. It is possible to develop an early maturity or short duration rice genotype using PHY-227 as a donor parent in breeding programs by using this as an early-flowering and mature genotype. The PHY-227 genotype showed the greatest mean performance for panicle length, followed by the PHY-226 genotype. The shortest and the tallest were PHY-228 and PHY-225, respectively. Grain yields per panicle were highest for genotype 238. Results similar to these are reported by Bekele et al. (2013). Genetic Variability: All the traits studied in this study had a PCV that was higher than their respective GCVs (Table 2). The PCV and GCV values differ very little for almost all the traits considered, indicating that environment is less of a factor affecting expression. Estimates of genetic parameters suggest that phenotypic differences may be taken into account as genotypic differences when selecting genotypes. The results were similar to those seen by Lingaiah et al. (2015); Mishu et al. (2016). PCV and GCV were highest for grain weight per plant, confirming that additive gene action is involved in determining these traits. Lal and Chauhan (2011); Singh et al. (2011); Singh et al. (2013) and Lingaiah (2015) also reported similar findings. A low PCV and GCV estimate showed that selection for traits such as the days to 50% flowering, the days to first flowering, the days to panicle initiation and kernel length and kernel breadth would be ineffective. It can be used to compare variability and sometimes provide clues on the validity of a trait for selection. For grain weight per plant, the PCV ranged from 5.915 % days to 50% panicle initiation, whereas for panicle length, the GCV ranged from 5.452 % to 24.252 %. The PCV and GCV values over 20% are considered high according to Sivasubramanian and Menon (1973), while values less than 10% are considered to be low and values between 10 and 20% are considered moderate. According to the delineated values, PCVs were high for panicles per plant, grain weight per plant, medium for chlorophyll, plant height, total number of tillers, effective number of tillers, kernel length to breadth ratio, and low for days to first panicle, days to first flowering, days to 50% flowering, panicle length, kernel length, kernel breadth. The highest genotype and phenotypic co-efficient of variation was obtained for grain weight per plant, followed by grain per panicle, number of effective tillers, and chlorophyll content. The lowest genotype and phenotypic co-efficient of variation was obtained for grains per panicle, and the highest were obtained for the number of effective tillers in both years. The highest number of grains per panicle as well as the highest number of tillers per plant was reported by Singh et al. (2006); Akhtar el al. (2011). According to Adhikari et al. (2018), grain yield per plant, number of panicles per plant, and number of fillers per plant were all associated with high genotypic and phenotypic coefficients of variation. The variance analysis found that all characters had significant differences between genotypes. Most of the traits under study displayed a wide range of variability. Estimates of phenotypic (δ2p) and genotypic variances (δ2g), phenotypic coefficient variance (PCV) and genotypic coefficient of variance (GCV), heritability (%), and genetic advance (GA) are shown in Table 2. With chlorophyll contents of 1044.61 and 258,772, respectively, there was a high degree of phenotypic and genotypic variance. High phenotypic and genotypic variances were equally observed for grains per panicle (693.356 and 103.223), plant height (271.062 and 112.492 ), grain weight per plant (24.796 and 17.455), days to first flowering 19.925 and 66.302), days to first panicle initiation (17.871 and 58.197), days to 50% flowering (17.787 and 71.302), panicle length (2.930 and 24.815), total number of tillers (0.828 and 8,502), seed index (0.127 and 2.610), kernel length (0.285 and 6.489) kernel breadth (0.038 and 2.093), kernel length- breadth ratio (0.130 and 3.122) grain weight per plant (37.244 and 24.252), respectively. Across all characters, phenotypic variance was higher than genotypic variance. Correlation: In order for grain yield to be expressed, a variety of factors need to be taken into account. Grain yield is controlled by minor genes. Selection for yield in and of itself isn't always effective. The inclusion of yield-attributing traits which show positive and significant associations with grain yield is therefore highly desirable for improving grain yield. An analysis of correlations revealed a significant positive correlation between days to panicle initiation and days to 50% blooming over both years. The days from panicle initiation to plant height correlated negatively and significantly. On the other hand, plant height did not show a significant correlation with days to first flower. Chrorophyll content and seed index correlated positively with days to first flowering. Kernel length and seed index showed a strong positive relationship over time. As well, kernel length and kernel length-breadth ratios were positively correlated over years, whereas kernel breadth was negatively correlated with kernel length-breadth ratio. Similarly, Satyanarayana et al. (2005); Ramakrishnan et al. (2006) reported similar findings. Thus, it is necessary to understand the nature of the relationship between yield and its components during crop improvement programmes so that the character of the crop can be simultaneously enhanced. Heritability. Heritability estimates were high for days to panicle initiation (95.6%), days to flowering at 50% (94.6%), and plant height (91.6%). On the other hand, the lowest heritability estimate was recorded for number of effective tillers (39.9), total number of tillers (46.1%), kernel breadth (48.5%), kernel length (54.6%), and kernel length-breadth ratio (57.9%). Similar studies have also been published by Singh et al. (2007), Fiyaz et al. (2011), Singh et al. (2013), and Islam et al. (2016). These results indicate that these traits can be effectively selected compared to others. Gene action can occur both additively and non-additivley, based on heritability estimates. High heritability is not necessarily indicative of high genetic gain. It is more efficient to use both heritability and genetic advance as selection parameters rather than the heritability alone, since both indicate that the heritability is most likely the result of additive gene effects.