Background: The excruciating pain of patients with renal colic on presentation to the emergency department requires effective analgesia to be administered in the shortest possible time. Trials comparing intramuscular non-steroidal anti-inflammatory drugs with intravenous opioids or paracetamol have been inconclusive because of the challenges associated with concealment of randomisation, small sample size, differences in outcome measures, and inadequate masking of participants and assessors. We did this trial to develop definitive evidence regarding the choice of initial analgesia and route of administration in participants presenting with renal colic to the emergency department. Methods: In this three-treatment group, double-blind, randomised controlled trial, adult participants (aged 18-65 years) presenting to the emergency department of an academic, tertiary care hospital in Qatar, with moderate to severe renal colic (Numerical pain Rating Scale ≥4) were recruited. With the use of computer-generated block randomisation (block sizes of six and nine), participants were assigned (1:1:1) to receive diclofenac (75 mg/3 mL intramuscular), morphine (0·1 mg/kg intravenous), or paracetamol (1 g/100 mL intravenous). Participants, clinicians, and trial personnel were masked to treatment assignment. The primary outcome was the proportion of participants achieving at least a 50% reduction in initial pain score at 30 min after analgesia, assessed by intention-to-treat analysis and per-protocol analysis, which included patients where a calculus in the urinary tract was detected with imaging. This trial is registered with ClinicalTrials.gov, number NCT02187614. Findings: Between Aug 5, 2014, and March 15, 2015, we randomly assigned 1645 participants, of whom 1644 were included in the intention-to-treat analysis (547 in the diclofenac group, 548 in the paracetemol group, and 549 in the morphine group). Ureteric calculi were detected in 1316 patients, who were analysed as the per-protocol population (438 in the diclofenac group, 435 in the paracetemol group, and 443 in the morphine group). The primary outcome was achieved in 371 (68%) patients in the diclofenac group, 364 (66%) in the paracetamol group, and 335 (61%) in the morphine group in the intention-to-treat population. Compared to morphine, diclofenac was significantly more effective in achieving the primary outcome (odds ratio [OR] 1·35, 95% CI 1·05-1·73, p=0·0187), whereas no difference was detected in the effectiveness of morphine compared with intravenous paracetamol (1·26, 0·99-1·62, p=0·0629). In the per-protocol population, diclofenac (OR 1·49, 95% CI 1·13-1·97, p=0·0046) and paracetamol (1·40, 1·06-1·85, p=0·0166) were more effective than morphine in achieving the primary outcome. Acute adverse events in the morphine group occurred in 19 (3%) participants. Significantly lower numbers of adverse events were recorded in the diclofenac group (7 [1%] participants, OR 0·31, 95% CI 0·12-0·78, p=0·0088) and paracetamol group (7 [1%] participants, 0·36, 0·15-0·87, p=0·0175) than in the morphine group. During the 2 week follow-up, no additional adverse events were noted in any group. Interpretation: Intramuscular non-steroidal anti-inflammatory drugs offer the most effective sustained analgesia for renal colic in the emergency department and seem to have fewer side-effects. Funding: Hamad Medical Corporation Medical Research Center, Doha, Qatar.