The increasing use of trophic position (TP) for assessing and describing ecosystems has resulted in the emergence of novel techniques for accurate and accessible measurements. The strength of amino acid compound specific nitrogen isotopic analysis (AA‐CSIA) lies in its ability to determine TP using a single organism, utilizing the isotopic differences between two or more amino acids (AAs). However, calculating TP from this technique relies upon the predictability of isotopic differences across consumer types and consistency among ecosystems. The goal of this study was to use wild samples to evaluate 15N enrichment between AA groups with differing trophic fractionation patterns, termed the trophic discrimination factor (TDFAA), and determine an accurate value for calculating TPs of marine teleosts. Using a large sample size (224 samples of 47 species) across a broad range of TPs (2.0–4.5, estimated by stomach content analysis [SCA]) and marine ecosystems, we derived TDFAA values that explained up to 80% of the variability between TP calculations from SCA and AA‐CSIA. We found a TDFAA value of 5.7 ± 0.3‰ using the difference between δ15N values of glutamic acid and phenylalanine, 2‰ lower than the most commonly used value of 7.6 ± 1.3‰. The weighted mean of several AAs resulted in the same value for TDFAA, but provided greater agreement with TPs from SCA. For more reliable TP calculations in marine teleosts, our results strongly advocate a lower TDFAA and the use of weighted mean AA δ15N values from triplicate analyses.