In this paper we propose guidelines for clinical trials aimed at assessing the efficacy of drugs for acute non-specific low back pain (LBP) with or without radicular pain, preliminary to their approval and registration. To this end, consensus statements were obtained from a group of experts in the fields of rheumatology, clinical medicine, public health and epidemiology. EBM resources were systematically used as references. Four diagnostic categories were defined: type 1--LBP with no radiation; type 2--LBP radiating no further than the knee; type 3--LBP radiating beyond the knee, but with no neurologic signs; and type 4--LBP radiating to a specific and entire leg dermatome, with or without neurologic signs. Studies should be designed on the basis of the claimed indications for the drug, but must be double-blinded whatever the indication. The duration of the study may be shorter for LBP type 1 or 2 (one week) than for LBP types 3 and 4 (up to one month), depending on the aim of the study and the indications for the drug. The comparator may be inactive (placebo) or active (for a superiority trial, e.g., versus paracetamol). Specific inclusion and exclusion criteria have been defined here for each category. An appropriate wash-out period for any drugs that could affect the pain status should be planned. Paracetamol may be allowed as rescue medication. The primary endpoint should be based on a validated pain assessment tool that may be either generic (type 1 or 2) or oriented (back and knee for types 3 and 4). Secondary endpoints could include the assessment of functional performance; the duration of any period of bed-rest; work limitation; a global assessment comprising pain at rest, standing and walking; the time elapsed before epidural injection, the prescription of other therapeutic agents, or surgery; and the use of rescue medication. Adverse events (AE) should be monitored systematically using a methodology that reflects the mode of action of the tested drug. With the application of these guidelines, LBP could serve as an appropriate disease for testing analgesic drugs. Rigorous evaluation may also help to improve the management of acute LBP.