The Qingyuan partridge chicken, a highly valued indigenous breed in China, requires optimised egg production traits to maximise economic returns. This study analysed weekly egg production data from 6 776 Qingyuan partridge chickens to compare the Grossman and Wood models in fitting individual egg-laying curves and identify the most robust model for estimating egg production persistency. Genetic parameters were estimated for three key traits: age at first egg (AFE), cumulative egg number up to 28 weeks (EN), and persistency derived from the Wood model, alongside weekly egg numbers using a random regression model. Results demonstrated the Wood model's superior performance, achieving successful curve fits for more individuals with comparable residual SEs. Moderate heritability estimates were observed for AFE (0.37 ± 0.02) and EN (0.28 ± 0.02), while persistency exhibited lower heritability (0.09 ± 0.06). Moderate to high genetic correlations emerged between AFE and EN (-0.63 ± 0.002), AFE and persistency (-0.51 ± 0.008), and EN and persistency (0.58 ± 0.008). The heritability of weekly egg numbers from weeks 1 to 28 followed a W-shaped trajectory, ranging from 0.07 to 0.19. Notably, genetic correlations between weekly egg numbers at weeks 16-19 and EN exceeded 0.91. These findings highlight the importance of selecting for egg production persistency alongside AFE and EN. Selection based on weekly egg numbers during weeks 16-19 offers a practical alternative to selection based on cumulative egg numbers, enabling early selection in breeding programmes.