Advice and exercise are widely recommended for subacute low back pain, but the effectiveness of these interventions is unclear. To investigate the effectiveness of physiotherapist-prescribed exercise, advice, or both for subacute low back pain. Factorial randomized, placebo-controlled trial. 7 university hospitals and primary care clinics in Australia and New Zealand. 259 persons with subacute low back pain (>6 weeks and <3 months in duration). Participants received 12 physiotherapist-directed exercise or sham exercise sessions and 3 physiotherapist-directed advice or sham advice sessions over 6 weeks. Primary outcomes were average pain over the past week (scale, 0 to 10), function (Patient-Specific Functional Scale), and global perceived effect (11-point scale) at 6 weeks and 12 months. Secondary outcomes were disability (Roland-Morris Disability Questionnaire), number of health care contacts, and depression (Depression Anxiety Stress Scales-21). Exercise and advice were each slightly more effective than placebo at 6 weeks but not at 12 months. The effect of advice on the pain scale was -0.7 point (95% CI, -1.2 to -0.2 point; P = 0.011) at 6 weeks and -0.4 point (CI, -1.0 to 0.3 point; P = 0.27) at 12 months, whereras the effect of exercise was -0.8 point (CI, -1.3 to -0.3 point; P = 0.004) at 6 weeks and -0.5 point (CI, -1.1 to 0.2 point; P = 0.14) at 12 months. The effect of advice on the function scale was 0.7 point (CI, 0.1 to 1.3 points; P = 0.014) at 6 weeks and 0.6 point (CI, 0.1 to 1.2 points; P = 0.023) at 12 months, and the effect of exercise was 0.4 point (CI, -0.2 to 1.0 point; P = 0.174) at 6 weeks and 0.5 point (CI, -0.1 to 1.0 point; P = 0.094) at 12 months. The effect of advice on the global perceived effect scale was 0.8 point (CI, 0.3 to 1.2 points; P < 0.001) at 6 weeks and 0.3 point (CI, -0.2 to 0.9 point; P = 0.24) at 12 months, and the effect of exercise was 0.5 point (CI, 0.1 to 1.0 point; P = 0.017) at 6 weeks and 0.4 point (CI, -0.1 to 1.0 point; P = 0.134) at 12 months. When administered together, exercise and advice had larger effects on all outcomes at 6 weeks (effect on pain, -1.5 [CI -2.2 to -0.7 point; P = 0.001], with similar results for other primary outcomes); however, by 12 months, there was a statistically significant effect only for function (effect, 1.1 points [CI, 0.3 to 1.8 points]; P = 0.005). Physiotherapists were not blinded. In participants with subacute low back pain, physiotherapist-directed exercise and advice were each slightly more effective than placebo at 6 weeks. The effect was greatest when the interventions were combined. At 12 months, the only effect that persisted was a small effect on participant-reported function. AUSTRALIAN CLINICAL TRIALS REGISTRY REGISTRATION NUMBER: 12605000039684.