In aquatic systems, the ratio of predator mass to prey mass (PPMR) is an important constraint on food web structure, and has been correlated with environmental stability. One common approach of estimating PPMR uses nitrogen stable isotopes (δ15N) as an indicator of trophic position, under the assumption that the discrimination between diet and tissue is constant with increasing diet δ15N (an additive approach). However, recent studies have shown that this assumption may not be valid and that there is a negative trend between the δ15N of the diet and the discrimination value (a scaled approach). Here, we estimated PPMR for a simulated food web using both the traditional additive approach and the improved scaled approach, and then tested our predictions with isotope samples from a North Sea food web. Our simulations show that the additive approach yields incorrect estimates of PPMR, and these biases are reflected in North Sea PPMR estimates. The extent of the bias is dependent on the baseline δ15N and trophic level sampled, with the greatest differences for samples with low baseline δ15N sampled at lower trophic levels. The scaled approach allows for the comparison of PPMR across varying δ15N baselines and trophic levels, and will refine estimates of PPMR.