The objective of this cohort study was to determine the association between the use of tramadol in emergency departments and the later consumption of opioids at the outpatient level in a group of patients from Colombia. Based on a medication dispensation database, patients over 18 years of age treated in different clinics in Colombia who for the first time received tramadol, dipyrone, or a nonsteroidal anti-inflammatory drug (NSAID) in the emergency room between January and December 2018 were identified. Three mutually exclusive cohorts were created, and each patient was followed up for 12 months after the administration of the analgesic to identify new formulations of any opioid. A Cox proportional-hazards regression model was constructed to identify variables associated with receiving a new opioid. A total of 12,783 patients were identified: 6020 treated with dipyrone, 5309 treated with NSAIDs, and 1454 treated with tramadol. The mean age was 47.1 ± 20.4 years, and 61.6% were women. A total of 17.3% (n = 2207) of all patients received an opioid during follow-up. Those treated with tramadol received a new opioid with a higher frequency (n = 346, 23.8%) than the other cohorts (14.7% NSAIDs and 17.9% dipyrone, both <math xmlns="http://www.w3.org/1998/Math/MathML" id="M1"> <mi>p</mi> <mo>&lt;</mo> <mn>0.001</mn> </math> ). In the tramadol group, using more than 10 mg of morphine equivalents was associated with a greater use of new opioids (HR:1.47, 95%CI:1.12–1.93). Patients treated with tramadol in emergency departments have a higher risk of opioid use at the one-year follow-up than those treated with NSAIDs or dipyrone.