INTRODUCTION Tomato (Solanum esculentum L.) bags the second position in the world among the most consumed vegetable after potato with a pro¬duction of 20573 thousand tonnes from 812 thousand ha area (NHB, 2020). Tomato is number one processing vegetable and is also popular due to its high vitamin A, vitamin C, potassium, phosphorous, magnesium and calcium contents. It is also rich in lycopene and beta-carotene, an antioxidant which promotes good health. Tomato finds various uses like for fresh market, processing, and also is one among the most preferred crop for protected cultivation, the crop being highly self-pollinated due to chasmogamy. Determinate growth habit is preferred for the processing tomatoes, whereas protected cultivation requires indeterminate varieties. Hence the objective of crop improvement in tomato is ever changing such as high yield, earliness and pest and disease resistance, improved fruit quality traits etc. Characterization and diversity assessment are two pillars important for utilizing distinctive cultivars in breeding programs and in conserving the genetic resources (Sridhar et al., 2022). A thorough study and evaluation of available germplasm is of great significance for both genetic and agronomic improvement of the crop (Reddy et al., 2013). The genetic variability and diversity existing can be evaluated and recorded using morphological, biochemical and molecular markers, morphological being the most budget friendly and accessible one. The present study focused on the aim of characterizing and assessing the variability of tomato accessions from different breeding environments. MATERIALS AND METHODS The study was conducted at Department of Vegetable Science, College of Agriculture, Thrissur, during 2020-2021. The experimental material comprised of exotic lines collected from World Vegetable Center, exotic lines collected from NBPGR, New Delhi, varieties and improved lines of Kerala Agricultural University. The experiment was laid out in randomized block design (RBD) with two replications with 20 plants per plot. All crop management practices were undertaken as per the Package of Practices Recommendations–Crops, Kerala Agricultural University, (2016). The data were recorded from five plants per replication on qualitative and quantita¬tive characters following the International Plant Genetic Resources Institute (IPGRI) descriptors for tomato (Table 1). Data on plant height, days to flowering, days to harvest, fruit length, fruit width, fruit weight, number of locules per fruit and TSS. Coefficients of variation were calculated as per Comstock and Robinson (1952). Heritability in broad sense and expected genetic advance were worked out as per Al¬lard (1961) and Johnson et al. (1955), respectively. RESULTS AND DISCUSSION Characterization based on 9 quantitative and 18 qualitative characters revealed significant variation among the genotypes (Table 2). 18 genotypes recorded determinate growth habit while, 7 were semi determinate. Similar difference in growth habit was reported by Agarwal et al. (2014). Leaf types varied from potato leaf, standard and peruvianum types. Potato leaves and standard leaves were observed in seven accessions each while 4 accessions was grouped under peruvianum type (Fig. 1). Presence of these three types of leaves were highlighted by Bhattarai et al. (2018). Foliage density varied among accessions. 5 accessions recorded dense foliage, 2 recorded sparse foliage and the remaining accessions had intermediate foliage (Fig. 7). All the accessions studied had yellow coloured perfect flowers with an open corolla blossom type. These results are in line with Salim et al. (2018). Fruit descriptors were more promising to be utilized in differentiating cultivars (Arvindkumar et al., 2003) Fruit shape and size play an important role in consumer acceptability of tomato. A good shaped fruit also ensures better packaging of the fruits and optimizes the space requirement. Besides, Fruit shape could be easily noticed and utilized for identifying tomato cultivars. In the present study, the shape of fruits ranged from round to highly rounded, flat to slightly flattened. Two accessions produced oblate or flattened fruits, whereas, slightly flattened fruits and round fruits were observed in 7 accessions. Three genotypes produced round fruits while high round fruit were observed in six accessions. The results agreed with the findings of Bhattarai et al. (2018), who gave reports of various fruit shapes like flattened, slightly flattened, cylindrical, rounded, high-rounded, and heart-shaped. The accessions also showed variation in the exterior colour of the immature fruits. Nine genotypes were observed with greenish-white fruits at an immature stage, while seven produced light green and one produced green (Fig. 2). This variation in colour may be due to genotypic variation among the accessions and various environmental factors (Salim et al., 2018). Similar color variation in fruits at immature stage were reported by Bhattarai et al. (2018) where he reported the existence of greenish-white, green, light green, dark green, very dark green and dark color fruits. Mature fruit color was red for the fifteen accessions studied, while three gave orange coloured fruits. The results are in accordance with Parisi et al. (2016) who found three colour variations namely orange pink and red in tomato. Two among the studied accessions produced fruits with prominent green shoulders (Fig. 3). This is in line with the reports of Parisi et al. (2016); Sacco et al. (2015) where they emphasized the presence or absence of green shoulder in tomato. Fruit shoulder shape varied from flat, slightly depressed and moderately depressed. Four accessions produced fruits with flat shoulder, whereas, five accessions gave fruits of moderately depressed and nine showed slightly depressed fruit shoulder. These results agreed with the findings of Figas et al. (2014). Similar results were observed by Bhattarai et al. (2018) such as flat, slightly flat, notched, slightly-depressed, moderately-depressed, depressed and strongly-depressed shoulder shape. Fruit cross-sectional shape was categorized into rounded and irregular shape in the study. All accessions produced fruits with round cross section whereas, LE19 was unique in producing angular shaped fruits. All the four types of pistil scars namely dot, stellate, linear and irregular were observed (Fig. 4). This is in line with the findings of Terzopoulos and Bebeli (2010) who reported different cross sectional shapes and the presence of four types of pistil scar in tomato. Flat blossom end was observed in all the eighteen genotypes. Similar observation of predominant flat blossom end fruits was made by Salim et al. (2018). This character is a more stable one and is less effected by various biotic and abiotic stresses and is most reliable character for cultivar differentiation in tomato (Vishwanath et al., 2014). Among the studied, two accessions namely LE 32 and LE 22 exhibited fruit cracking – concentric and radial cracking respectively (Fig. 5, 6). These findings are in consonance with results of Figas et al. (2014) who reported radial and concentric cracking in tomato. The analysis of variance revealed significant differences among the genotypes with respect to all the characters. The extent of variability in range, mean, genotypic co-efficient of variance (GCV), phenotypic co-efficient of variance (PCV), heritability, expected genetic advance and the expected genetic advance as per cent of mean are presented in Table 3. Plant height, fruit width and fruit weight showed high values for PCV, which indicated the influence of environment on these when compared to other characters. Majority of the characters showed moderate PCV and GCV whereas, fruit weight and yield per plant experienced a high GCV and PCV. This agreed with Patel et al. (2013) where they obtained moderate values for plant height, days to flowering, days to harvest and locules per fruit. High GCV and PCV for fruit weight, fruit yield per plant and moderate to high GCV and PCV was observed in fruit length and fruit width was reported in a study by Singh et al. (2019). The relative magnitude of difference between PCV and GCV was low for all characters except plant height indicating the low influence of environmental factors on these characters. These findings suggested that selection on phenotypic basis is effective along with equal probability of genotypic values. GCV alone, cannot be used to conclude the extent of heritable variation and hence, the knowledge of heritability too is needed for selection. The genetic advance for quantitative characters helps in achieving selection. Highest heritability (broad sense) were found for average fruit weight (98.7%), fruit yield per plant followed by fruit width (82.9%), number of locules (81.5 %), fruit length (80.1 %) and TSS (79.2 %), as observed by Singh et al. (2019); Prakash et al. (2019), whereas, plant height and days to harvest exercised moderate heritability. Days to flowering recorded the lowest value for heritability at broad sense, which is in line with the results of Kerketta and Bahadur, (2019). Genetic advance as per cent of mean was highest for all characters except days to flowering and days to harvest. Highest estimates of heritability along with high genetic advance were observed for average fruit weight, fruit width, fruit length, TSS and number of locules. The above findings agreed with Tasisa et al. (2011).