This study investigates biometric relationships in Oreochromis niloticus (gray tilapia) reared in controlled pond environments at the Centro de Investigación Piscícola (CINPIS), Universidad Nacional Agraria La Molina (Perú), over a 3-year period (2021–2023). Focusing on total length ( \({L}_{t}\) ), standard length ( \({L}_{s}\) ), height ( \(H\) ), and width ( \(A\) ), we developed models to estimate weight ( \(W\) ) based on these parameters, achieving strong model performances with R2 values between 0.899 and 0.994. The model using \({L}_{t}\) as a predictor of \(W\) proved most accurate with a mean relative error (MRE) of 11.2%, while models incorporating additional dimensions ( \(H\) and \(A\) ) did not enhance predictive accuracy. Comparative analyses show our model aligns with some studies on tilapia, though variations in L-W relationships due to environmental and breeding conditions are evident. Our results affirm the utility of \({L}_{t}\) in non-invasive biomass estimation for aquaculture, while highlighting the limitations of applying these models universally across different conditions and fish populations. Condition factor ( \(K\) ) and relative condition factor ( \({K}_{r}\) ) analyses further demonstrated stable and optimal growth conditions (mean \(K\) ≈ 1.76; \({K}_{r}\) ≈ 1.01) for tilapia under local culture practices. Accordingly, we propose the L-W relationship \(W= {0.0265L}^{2.8469}\) for estimating the weight of gray tilapia grown in ponds with groundwater. This study provides a basis for the development of biomass estimation methods based on active acoustics or stereo video.
