© 2017, © 2017 Informa UK Limited, trading as Taylor & Francis Group. Purpose: To establish the within-day and between-day intra-rater reliability, inter-rater reliability, validity and systematic errors of the tandem gait test (TGT). Materials and methods: Thirty participants performed the TGT and the timed up and go test (TUG) twice on the first day. Three independent raters measured these tests. After two weeks, the protocol was repeated by one tester to assess between-day reliability. The within-day and between-day intra-rater reliability was estimated by intraclass correlation coefficient (ICC) and Bland–Altman plots to assess systematic errors associated with the learning effect. Inter-rater reliability was estimated by ICC and minimum detectable change at 95% confidence. Concurrent validity was assessed by examination of the correlation between TGT and TUG using Pearson’s correlation coefficient. Results: Within-day and between-day intra-rater reliability were good (ICC > 0.9 and ICC > 0.7). The Bland–Altman plots showed systematic error in the between-day measurements. Inter-rater reliability was supported by strong values for ICC (ICC > 0.9). The Pearson’s correlation coefficient between the TGT and TUG showed a significant positive correlation (r > 0.67, p <.001). Conclusions: These results show that the learning effect of the TGT might be a considerable factor when interpreting follow-up measurements after a few days.