Variability, correlation and path coefficient analysis were undertaken in F4 populations of okra (Abelmoschus esculentus L. Moench) derived from the cross OK-2017-010 × Pusa Sawani during Summer 2025 at the Department of Biotechnology and Crop Improvement (BCI), College of Horticulture, Bengaluru. A total of 11 families comprising 220 plants, two parents and five standard checks were evaluated in an Augmented Block Design for 14 yield and yield-contributing traits. High broad-sense heritability (h²bs) coupled with high genetic advance as percent mean (GAM) was recorded for total yield per plant (91.70%; 36.72%), days to 50% flowering (92.50%; 36.92%), number of fruits per plant (90.90%; 31.41%) and fruit diameter (86.80%; 35.66%), indicating the preponderance of additive gene action and the feasibility of direct selection for these traits. Number of fruits per plant exhibited the highest positive genotypic coefficient of variation (GCV = 25.99%) and phenotypic coefficient of variation (PCV = 26.77%), followed by total yield per plant (GCV = 28.61%; PCV = 29.44%). Correlation analysis revealed that number of fruits per plant (r = 0.898**) and number of branches per plant (r = 0.844**) had the strongest positive association with total yield per plant, while days to first flowering, plant height, fruit diameter, fruit length and number of seeds per fruit showed significant negative correlations with yield. Path coefficient analysis identified number of fruits per plant as the primary direct contributor to total yield (direct effect = 1.045), followed by number of seeds per fruit (0.649) and number of branches per plant (-0.145 with high indirect effects). These findings provide valuable insights for the improvement of okra yield through recurrent selection and family-based breeding in segregating populations.