INTRODUCTION Brinjal (Solanum melongena L., 2n=24) belongs to the Angiospermic family ‘Solanaceae’ and it is an often-cross pollinated crop with cross pollination reported as high as 48% (Madhavi et al., 2015). China is the world’s leading eggplant producer with over half of eggplant acreage, followed by India which accounts for roughly one quarter of total world’s production. Overall, Asia accounts for about 94 percent of the world eggplant area with about 92 percent of world output (FAO, 2010). The eggplant is well adapted to grow under high rainfall and high temperatures, as well as under dry conditions with irrigation. Eggplant has moderate amounts of dietary fibre, vitamins, and micronutrients and it contributes to the diet of people in developing countries when other vegetables are in short supply. Due to its low calorie (24kcal/100g) and high potassium content (200mg/100g), it is suitable for diabetes, hypersensitive and obese patients. There are umpteen number of commercially grown varieties and hybrids available in the market released by both public and private sector. However, a genotype possessing considerably high yield potential coupled with stable performance in different environments has great value for its adaptation on large scale and in plant breeding programme (Mehta et al., 2011; Raj et al., 2019). Moreover, there is an utmost need for development of high yielding stable varieties and hybrids for specific environments and seasons (Vadodaria et al., 2009). Genotype and environmental interaction play a significant role for any such productive gain. Selection of suitable and stable crop varieties has received much attention by the breeders as an advance approach in increasing crop production. A stable variety/hybrid is desirable for obtaining uniform crop yield over a wide range of agro-climatic situations. Stability in productivity is a major and it is important to identify brinjal genotypes capable of performing well across the environments (Sofiya & Raj, 2021). Study of stability parameters is useful to measure adaptability and stability of crop cultivars, which can be used to identify genotypes suitable for different environments from season to season. Genotype × Environment interaction is expected to play an important role in the performance of genotypes under diverse environmental conditions, besides their individual effect. Among various other stability models, Eberhart and Russell (1966) model is the predominantly used one. Krishna et al. (2022) studied mango cultivars and found Mallika to be the most stable variety using Eberhart and Russell model. Raj et al. (2019) utilized the same model and identified AU-101 as a stable hybrid under unfavourable conditions. Sara et al. (2021) studied thirty-three pearl millet genotypes for stability using the same model- and found AUBH-15 to be stable hybrid for yield. Mehta et al. (2011) studied seven long brinjal varieties and found that IBWI-2007-1 to be stable under irrigated conditions. Sivakumar et al. (2017) studied thirty four brinjal genotypes and found four brinjal hybrids to be stable for fruit yield per plant. Studied fifty five brinjal genotypes identified Pusa Uttam and Pusaupkar to be stable over four different environments. Siva et al. (2020) studied a total of thirty brinjal genotypes and found four hybrids to be stable across environment for fruit yield and its components. The regression model of stability has been widely used in many annual crops; however, the literature is still limited for brinjal. Also, in a geographical point of view, the available studies on brinjal are mainly conducted in North India and more assessment of environmental effects on brinjal fruit yield under South Indian conditions is also a necessity. Considering the above points this study was designed to study the brinjal genotypes thar are having stable performance for yield and its components throughout the year in different environments. MATERIALS AND METHODS The present investigation was conducted in three different locations viz., experimental farm in the Department of Genetics and Plant Breeding, garden land farm in the Department of Agronomy of Faculty of Agriculture, Annamalai University, Annamalai Nagar, Tamil Nadu and experimental farm in Virudhachalam, Tamil Nadu. The experimental materials for this study comprised of thirty-six brinjal genotypes obtained from the National bureau of plant genetic resources (NBPGR) and some land races and released varieties (Table 1). Seeds of the 36 genotypes and land races were sown in raised nursery beds. The seedings are ready for transplanting within 4-6 weeks of planting when they attain a height of 15cm with 2-3 true leaves with the spacing of 60cm between rows and 60cm between plants. The experiment was carried out in randomized block design with three replications. Fifteen plants per replication were maintained for each genotype. Recommended agronomic practices and need based plant production measures were carried out. The observations were recorded on five traits viz., fruit length, fruit girth, number of fruits per plant, average fruit weight and fruit yield per plant. The data was subjected to Eberhart and Russell (1966) model of stability analysis and the phenotypic stability was determined by Lewis (1954) stability factor. The statistical analysis was carried out using TNAUSTAT software. RESULTS AND DISCUSSION The join regression analysis indicated that the variance due to genotype was significant for all the traits suggesting the presence of genetic variability among the genotypes under study (Table 2). The variances due to environment were also significant for all the five traits indicating that these traits were highly influenced by all the three locations. The variance due to G × E and E + (G × E) were also significant for all the traits inferring the differential response of the brinjal genotypes in different locations and when tested against pooled deviation it indicated that the genotypes differed widely among themselves. The E + (G × E) interaction was only significant for yield per plant against pooled error which indicated the distinct nature of seasons and G × E interactions in the phenotypic expression of the genotypes for this particular trait. High magnitude of the environment (linear) effect in comparison to Genotype × Environment (linear) for yield per plant was observed, which may be responsible for high variation to the trait in different locations. Similar observations for yield per plant was also observed by Chaitanya and Reddy (2021); Bora et al. (2011). Once the Genotype × Environmental interaction was found to be significant, the next step was to identify the stable genotypes that interacted the least with environments giving a near consistent performance. The genotype is found to be stable based on non-significant deviation from regression coefficient, mean values and a regression coefficient value equal to unity. Thus, depending on the character, a genotype with high or low mean (desirable mean depends on the character), unity in regression coefficient and non-significant deviation from regression were considered as widely adapted and stable genotype. Above and below average stability is based on greater than unity bi and lesser than unity bi, respectively. The deviation from regression were non-significant and regression coefficient were around the unity for genotype G5 and more than unity in G4, G8, G13, G30, G31, G34, G35 indicating above average stability, whereas greater than unity was observed in G1, G3, G10, G12, G15, G17, G20, G28, G33 indicating below average stability for fruit length (Table 3a and 3b). The genotype G28 were found to be above average stability due to regression coefficient more than unity for fruit girth and non-significant deviation from regression. For number of fruits per plant, the genotypes G17 and G28 exhibited non-significant deviation from regression and regression coefficient were found to be equal to unity indicating stable performance. For average fruit weight, the genotypes G5, G6, G10, G13, G14, G17, G21, G24, G25 had non- significant deviation from regression and equal to unity for regression coefficient indicating stable performance across environments. For fruit yield per plant, the deviation from regression were non-significant and the regression coefficient were found to be around the unity for the genotype G9, G19, G36 indicating their stable performance Similar studies were also conducted to identify stable genotypes for fruit yield per plant, average fruit weight and fruit length by Sivakumar et al. (2017); Siva et al. (2020); Chaudhari et al. (2015); Suneetha et al. (2006); Mehta et al. (2011) in brinjal. The criterion for identifying a genotype with less fluctuation due to environment in characters is by measuring the ratio between the high mean in any environment and the low mean in any environment. This is the simple measure of the phenotypic stability of a genotype. The stability factor nearing a ratio of 1.00 indicated the maximum phenotypic stability (Table 4). The genotypes G4, G10, G12, G17, G23, G26, G29, G34, G35 showed around unity stability factor for fruit yield per plant. The genotypes G5, G22, G28 with poor adaptability according to regression model showed around unity stability factor for three characters viz., fruit length, fruit girth, average fruit yield per plant. Almost all the genotypes showed maximum phenotypic stability for character fruit length.