BACKGROUND: Addressing prolonged workplace sitting is an identified priority. A national implementation trial of BeUpstanding™ - an online workplace intervention supporting teams of desk-based workers to sit less and move more - successfully reduced worker sitting time in a large sample of workers. However, it is unclear which strategies workers used to sit less and move more, how usage changed following intervention, and how this related to changes in work activity and sitting. METHODS: BeUpstanding collected data from staff before and after the 8-week champion-delivered intervention regarding self-reported work behaviours (percentage of worktime sitting and moving; percentage sitting in prolonged bouts) and the usage (0 = never to 4 = always/nearly always) of 21 sit less/move more strategies (13 emphasised 'move more'). Strategy usage during versus before intervention were compared via linear mixed models. Strategy usage (0-4) and changes (-4-4) were tested in relation to post-intervention behaviours and behaviour changes using linear mixed models. Interaction tests and conditional inference trees compared strategies in their effectiveness. RESULTS: Across 1614 staff (43.1 ± 11.3 years; 66% female) from 66 workplaces, the number of strategies used at least sometimes averaged 9.56 (SE = 0.19) pre-intervention. Strategy usage increased significantly (p < 0.05) in the number of strategies used (2.45 [95% CI: 2.18, 2.73], p < 0.001), mean strategy usage (0.37 [0.33, 0.41], p < 0.001), move-more strategy usage, and for every strategy except active travel. Every strategy was used by > 10% of staff following intervention. Strategy usage and changes were significantly associated with all behaviours and behaviour changes (all p < 0.01). There were significant differences in the strength of these associations between strategies (p < 0.05) and for move-more strategies versus other strategies (p < 0.05); however, no strategies were statistically counterproductive. Conditional inference trees identified various combinations of strategies whose usage predicted outcomes. CONCLUSIONS: All strategies showed acceptability (used by > 10%), most were modifiable (increased with intervention), and, to varying degrees, their usage was associated with work behaviours. Findings suggest all 21 strategies are suitable for Australian desk-based workers to select based on personal and contextual fit. Strategies most strongly linked with all behaviours or targeted behaviours (i.e., increasing movement) might be emphasised to enhance effectiveness. TRIAL REGISTRATION: ACTRN12617000682347. The trial was prospectively registered on the 12th May, 2017 (ACTRN12617000682347) before the soft launch online and last updated on the 11th June 2019, before the national implementation trial recruitment commenced (12th June, 2019).