BACKGROUND: Neighbourhood walkability is known to be positively associated with self-reported and device-based measures of overall physical activity. However, relations of walkability with specific active and sedentary behaviour patterns are not well understood. METHODS: We investigated cross-sectional associations of neighbourhood walkability with time spent stepping, standing, sitting, and their pattern metrics using data from 505 participants (mean age 59.2 years) from the AusDiab3 study. Neighbourhood walkability (a composite measure of residential, destination, and intersection densities) was calculated within 1 km street-network buffers around participants homes. Thigh-worn device data (activPAL, 7-day, 24 h/day protocol) were used to derive stepping, sitting and standing minutes per day and their pattern metrics. Two-level linear mixed models assessed relevant associations, adjusting for potential confounders. RESULTS: Higher walkability was associated with higher cadences (β [95% CI] = 0.12 [0.04-0.20]), moderate-to-vigorous physical activity (β [95% CI] = 0.17 [0.09-0.26]), longer stepping bouts (β [95% CI] = 0.18 [0.10-0.25]) and time in purposeful (≥ 2 min duration) walking (β [95% CI] = 0.21 [0.13-0.30]). There were no associations with total sitting time, standing time, or their associated pattern metrics. Total stepping time also had no associations, suggesting that participants in neighbourhoods with higher walkability may accumulate similar levels of stepping time to participants in lower walkability neighbourhoods, albeit with higher intensity and in longer bouts. CONCLUSIONS: By examining activity totals only, relevant walkability relationships may be masked. Further research is needed to understand whether walkability and other built environment attributes are associated with sedentary behaviour patterns, as well as light-intensity physical activities.