Products will be judged as weaker in intensity when presented in a series with stronger items and stronger in intensity when presented with weaker items. This is a context effect known as simple contrast. Four groups of panelists were tested for the appearance and persistence of simple contrast effects in judgments of sweetness intensity. One group received training with intensity references, 1 group received a reference standard in each test, 1 group received both training and intensity references, and 1 group received neither training nor any references. In spite of extensive training and the presence of reference standards, the contrast effects were persistent. Training with intensity standards, as sometimes done in descriptive analysis, may not be sufficient to protect against shifts in judgments when products are judged in different contexts.