Abstract Objectives Among patients with primary and relapse visceral leishmaniasis (VL) in eastern Sudan, we determined the proportion eligible for treatment with sodium stibogluconate and paromomycin (SSG/PM) and, of these, their demographic and clinical characteristics; initial treatment outcomes including adverse side effects requiring treatment discontinuation; treatment outcomes by 6 months; and risk factors associated with initial (slow responders) and late treatment failure (relapses and post‐kala‐azar dermal leishmaniasis, PKDL). Methods A retrospective cohort study in Tabarak Allah Hospital, Gedaref Province, eastern Sudan, from July 2011 to January 2014. Results Of 1252 individuals diagnosed with VL (1151 primary and 101 relapses), 65% were eligible for SSG/PM including 83% children, almost half of them malnourished and anaemic. About 4% of individuals discontinued treatment due to side effects; 0.7% died during treatment. Initial cure was achieved in 93% of 774 primary cases and 77% of 35 relapse cases ( P < 0.001). Among the 809 patients eligible for SSG/PM, 218 (27%) were lost to follow‐up. Outcomes by six months among the 591 patients with available follow‐up data were: definitive cure ( n = 506; 86%), relapse ( n = 38; 6%), treatment discontinuation ( n = 33; 6%), PKDL ( n = 7; 1%) and death ( n = 7; 1%). Among those completing a full course of SSG/PM, relapses and under‐fives were at significantly higher risk of early and late treatment failure, respectively. Conclusion Whether SSG/PM as a first‐line regimen is an undeniable progress compared to SSG monotherapy, it excluded a considerable proportion of VL patients due to drug safety concerns. We call for accelerated development of new drugs and treatment regimens to improve VL treatment in Sudan.