**Summary:** In this study we investigated the effect of two nights of sleep restriction on reinforcement learning, that is learning from trial-and-error. We found that sleep restriction did not affect the generalized learning, measured in a test phase, or the win-stay/lose-shift learning tendencies in a learning phase. However, from computational modeling of the learning rate during the learning phase we observed a slower speed of learning from negative feedback after sleep restriction that could be related to increased risk-seeking behavior found in other studies. **Publication:** Gerhardsson, A., Porada, D.K., Lundström, J.N., Axelsson, J., Schwarz, J. Does insufficient sleep affect how you learn from reward or punishment? Reinforcement learning after 2 nights of sleep restriction. J Sleep Res, 2020,. https://doi.org/10.1111/jsr.13236 **Data** https://doi.org/10.17045/sthlmuni.11955939.v1 **Errata:** *Table 1*: Data for one of the participants on one of the measure dates for variable Sleep Diary: Well rested was wrongly inputed as 0, which is the reason that the minimum value 0.5 is less than possible value 1.