Objective: We investigated the effectiveness of error-encouragement training and the influence of ability and personality attribute-treatment interactions using a dynamic decision-making task. Background: Error-encouragement training is said to be more effective than error-avoidance training because active exploration and error encouragement promote better learning and transfer. Past research, which is limited by confounded experimental designs involving simple tasks with little consideration of individual differences, provides evidence for the value of active exploration but less so for error encouragement. Method: We randomly allocated 164 participants to receive error-encouragement, error-avoidance, or no error-related instructions in training on ATC-lab, a dynamic computer simulated air traffic control task. Active exploration and task information were controlled, and ability and personality were assessed. Results: Error encouragement yielded better transfer performance than did error avoidance but was no better than the control training. Higher- rather than lower-ability trainees benefited from training with an error orientation (positive or negative), suggesting that learning from errors in training requires greater cognitive resources. Trainees higher in openness to experience and agreeableness performed more poorly with error-avoidance training than with error-encouragement and control training. Conclusion: Error-avoidance training had a detrimental effect especially for those less open to experience, lower in agreeableness, and of lower ability. The significant benefits of error-encouragement training reported in previous research were probably attributable to confounding factors, such as active exploration and differential task information. Application: Although errors and active exploration can be useful learning tools in training, trainers should be mindful that focusing on errors, whether positively or negatively, may not suit everyone.