A history of ever-smoking appears to be associated with a more severe disease phenotype in axial spondyloarthritis (axSpA). However, evidence is sparse for the effect of increased smoking exposure on disease outcomes or whether smoking reduction or cessation improves outcomes. The aim of this study was to explore whether a dose–response relationship exists between pack-years and disease activity and functional impairment in axSpA. Consecutive patients meeting ASAS criteria for axial SpA were recruited from a spondyloarthritis service. The associations between pack-years of smoking and: (1) disease activity (BASDAI/ASDAS), (2) spinal pain, (3) functional impairment (BASFI) and (4) inflammatory markers were explored using multivariable linear models, adjusted for age, gender and use of TNF inhibition (TNFi) therapy. Pack-years were categorised into four groups (<10, 11–20, 21–40, >40) and analysed with light smoking (<10) as reference. Two hundred and thirty-eight axSpA patients were recruited: 76% were male, mean age 46.4 years (SD ± 13.7), and 33% were treated with TNFi. One hundred and twelve patients reported history of ever-smoking with median pack-year 20 [IQR10-30]. Compared to light smokers, those with higher categories of smoking exposures had higher BASDAI (21–40 pack-years, β = 1.6 (95% CI 0.28, 2.95); >40, β = 2.6 (0.54, 3.56)), higher BASFI (21–40, β = 2.1 (0.42, 4.80); >40, β = 3.2 (0.76, 5.71)), and higher ASDAS (21–40, β = 0.82 (0.14, 1.51)). This cross-sectional study demonstrated that smoking is associated with increased axSpA severity markers in a dose–response manner. Particular effort should be made to restrict smoking exposure early before accruing a significant number of pack-years.