Objectives: The long-term side effects of stavudine (d4T) led to recommendations in 2009 to phase out use of this drug. We aimed to describe temporal patterns of d4T use across Europe. Methods: Patients taking combination antiretroviral therapy (cART) in EuroSIDA with follow-up after 1 January 2006 were included in the study. cART was defined as d4T-containing [d4T plus at least two other antiretrovirals (ARVs) from any class] or non-d4T-containing (at least three ARVs from any class, excluding d4T). Poisson regression was used to describe temporal changes in the prevalence of d4T use and factors associated with initiating d4T. Results: A total of 5850 patients receiving cART on 1 January 2006 were included in the current analysis, rising to 7768 patients on January 1 2013. During this time, the prevalence of d4T use fell from 11.2% to 0.7%, with an overall decline of 19% per 6 months [95% confidence interval (CI) 19-20%]. d4T use declined fastest in Northern Europe [26% (95% CI 23-29%) per 6 months], and slowest in Eastern Europe [17% (95% CI 16-19%) per 6 months]. In multivariable Poisson regression models, new d4T initiations decreased by 14% per 6 months [adjusted incidence rate ratio (aIRR) 0.86; 95% CI 0.80-0.91]. Factors associated with initiating d4T were residence in Eastern Europe (aIRR 4.31; 95% CI 2.17-9.98) versus other European regions and HIV RNA>400 copies/mL (aIRR 3.11; 95% CI 1.60-6.02) versus HIV RNA