BACKGROUND: To improve the effectiveness of gratitude exercises in reducing anxiety and worry, app-assisted variations of the 'Three Good Things' intervention, differing in timing and instructions, were conducted. METHOD: Healthy participants (n = 227) exercised either in the morning versus evening (writing down three things that induced or could induce gratitude during the day) or received identical versus varying instructions to guide the listings. Short-term intervention effects were recorded using a smartphone application before and after each daily exercise, while mid-term effects were assessed based on changes in questionnaire scores following the two-week intervention. RESULTS: Varied instructions had a greater impact on reducing state anxiety (worry and arousal). However, the timing of the exercises did not differentially influence the positive short-term and mid-term outcomes, including the reduction of state and trait worry. Participants preferred evening practice. CONCLUSION: Based on the observed practice effects and participants' stated preferences, evening exercises with varied instructions for the 'Three Good Things' intervention should be prioritized to counteract anxiety and worry.