The proportion of new HIV diagnoses between May and December 2009 across Odessa recently-infected was estimated using the BED-CEIA assay. Logistic regression models were used to explore factors associated with testing as recent. Of 1,313 newly-diagnosed individuals, 321 (24 %) were classified as recent. Recent infection was less likely among older adults [odds ratio (OR) = 0.70 per 10-year increase, 95 % CI 0.60–0.82]. Compared to men residing in Odessa city, women in rural Odessa and non-resident men were more likely to be recently-infected (OR 1.85, 1.26–2.71 and 2.83, 1.15–6.97, respectively). Reason for test was not associated with recent infection. In sensitivity analysis, after excluding individuals tested due to clinical indications, the proportion recently-infected and the association with age remained virtually unchanged. Our findings suggest a high risk of onward transmission, particularly in younger age groups. These findings highlight the need for tailored prevention strategies and ongoing RITA testing to monitor and evaluate effectiveness of prevention programmes.