Trained expert panels are used routinely in boar taint research, with varying protocols for training of panelists and scoring methods. We describe a standardized process for training and scoring, to contribute to standardize the olfactory detection of boar taint. Three experiments are described in which we (1) evaluate the importance of training and the effect of the previous sample, (2) determine detection thresholds on strips and in fat for our panel, and (3) test priming panelists before boar taint evaluation. For the final evaluation of boar taint, we propose a consistent three-person evaluation scoring on a 0–4 scale using a final mean score of 0.5 as the cut-off for boar taint. This gave an optimal sensitivity of 0.81 and a specificity of 0.56 compared to chemical cut-offs. Even limited training proved useful, but priming assessors with strips did not improve the evaluation of fat samples. Detection thresholds were higher in fat compared to strips, except for indole. We recommend panelists to always smell a non-tainted control sample after a tainted one as a ‘reset’ mechanism, before continuing. For longitudinal studies, we additionally advise to set up an expert panel with a fixed number of assessors performing each evaluation in duplicate.