<p>Red sea bream <i>Pagrus major</i> is one of the most important aquaculture species in Japan. Stereo cameras are widely used for monitoring fish growth. However, growth monitoring using stereo cameras encounters measurement issues, such as depth-related bias, which may be influenced by the fish’s depth distribution. Swimming speed is a key indicator of fish welfare and affects depth distribution. The objective of this study is to elucidate the spatiotemporal characteristics of red sea bream in aquaculture environments, with a particular focus on fork length and swimming speed. Experiments were conducted at the same site across four different seasons. Fork length and swimming speed were measured using four stereo cameras placed at different depths. The depth distributions were assessed using a single-beam sonar during summer and autumn. The results revealed a significant depth-related bias in fork length measurements throughout the year. Swimming speed varied with depth, season, and fish size. These findings suggest that the depth at which stereo cameras are installed is a critical factor influencing the accuracy of fork length measurements. In conclusion, the optimal depth for fork length measurement of red sea bream aquaculture is the middle depth of fish school.</p>