After implementation of a ‘fast-track’ rehabilitation protocol in our hospital, mean length of hospital stay for primary total hip arthroplasty decreased from 4.6 to 2.9 nights for unselected patients. However, despite this reduction there was still a wide range across the patients’ hospital duration. The purpose of this study was to identify which specific patient characteristics influence length of stay after successful implementation of a ‘fast-track’ rehabilitation protocol. A total of 477 patients (317 female and 160 male, mean age 71.0 years; 39.3 to 92.6, mean BMI 27.0 kg/m2;18.8 to 45.2) who underwent primary total hip arthroplasty between 1 February 2011 and 31 January 2013, were included in this retrospective cohort study. A length of stay greater than the median was considered as an increased duration. Logistic regression analyses were performed to identify potential factors associated with increased durations. Median length of stay was two nights (interquartile range 1), and the mean length of stay 2.9 nights (1 to 75). In all, 266 patients had a length of stay ≤ two nights. Age (odds ratio (OR) 2.46; 95% confidence intervals (CI) 1.72 to 3.51; p < 0.001), living situation (alonevsliving together with cohabitants, OR 2.09; 95% CI 1.33 to 3.30; p = 0.002) and approach (anterior approachvslateral, OR 0.29; 95% CI 0.19 to 0.46; p < 0.001) (posterolateral approachvslateral, OR 0.24; 95% CI 0.10 to 0.55; p < 0.001) were factors that were significantly associated with increased length of stay in the multivariable logistic regression model.