Improving the ability to predict persistent pain after spine surgery would allow identification of patients at risk and guide treatment decisions. Quantitative sensory tests (QST) are measures of altered pain processes, but in our previous study, preoperative QST did not predict pain and disability at single time-points. Trajectory analysis accounts for time-dependent patterns. We hypothesized that QST predict trajectories of pain and disability during 1 year after low back surgery. We performed a trajectory analysis on the cohort of our previous study (n = 141). Baseline QST included electrical, pressure, heat, and cold stimulation of the low back and lower extremity, temporal summation, and conditioned pain modulation. Pain intensity and Oswestry Disability Index were measured before, and 2, 6, and 12 months after surgery. Bivariate trajectories for pain and disability were computed using group-based trajectory models. Multivariable regressions were used to identify QST as predictors of trajectory groups, with sociodemographic, psychological, and clinical characteristics as covariates. Cold pain hypersensitivity at the leg, not being married, and long pain duration independently predicted worse recovery (complete-to-incomplete, incomplete-to-no recovery). Cold pain hypersensitivity increased the odds for worse recovery by 3.8 (95% confidence intervals 1.8-8.0, P < 0.001) and 3.0 (1.3-7.0, P = 0.012) in the univariable and multivariable analyses, respectively. Trajectory analysis, but not analysis at single time-points, identified cold pain hypersensitivity as strong predictor of worse recovery, supporting altered pain processes as predisposing factor for persisting pain and disability, and a broader use of trajectory analysis. Assessment of cold pain sensitivity may be a clinically applicable, prognostic test.