Species-specific length-weight relationships (LWRs) are valuable in fisheries science because fish size is often measured in terms of body length (Froese et al., 2014). For example, fish length estimates from non-destructive sampling methods (e.g., underwater visual census and remote underwater videos) are converted into weight using LWRs to provide fish biomass estimates (Daros et al., 2018;Wilson and Graham, 2018;Soeth et al., 2020). Moreover, detailed information on LWRs and their uncertainties allows us to investigate fisheries and their environmental impacts (Lehodey et al., 2008;Philippsen et al., 2019), to estimate the regional and global active carbon flux of fish (Saba et al., 2021), to estimate trophic interactions (Machado et al., 2020), and to calculate body condition indices and This study encompasses the description and evaluation of the length-weight relationship of 104 demersal fish species caught by bottom trawlers targeting shrimps on the southeast continental shelf of Brazil from 2004 to 2006. The regression criteria describing the length-weight relationship for each species were classified as approved (met the criteria), approved with reservations (partially met the criteria), and not approved (did not meet the criteria) based on linear regression parameters to determine whether length is a viable predictor of weight. A total of 141,433 individual fish, comprising 44 families and 104 species, were sampled; the beta parameter (± se) varied from 0.22 ± 0.12 to 3.94 ± 0.19, and the alpha parameter varied from -4.09 ± 0.04 to 0.89 ± 0.02. In total, 22 species were not identified by a recent large survey (2019) conducted in the study area. The results of this study are significant for the management of fishery resources, mainly due to the occurrence of unusual species, the economic importance and enormous effort exerted by the trawling fleet in the region, and the substantial sample size, in which a large number of individuals per species were caught.