Background: The peak in HIV viral load (VL) during primary HIV infection (PHI) may disproportionately contribute to the propagation of the epidemic. Whether there has been a temporal increase in VL during PHI is debatable as findings are conflicting. We examined whether VL at initial presentation and viral load set-point (VLSP), a proxy for virulence, has changed over time. Methodology: We used data from the UK Register of HIV Seroconverters and restricted analysis to men reporting HIV exposure through sex between men (MSM), diagnosed ≥01/01/1997 with a SC test interval (interval between negative and positive test dates) ≤12 months, and with ≥1 VL in the first 12 months. VLSP was defined as the mean of ART-naive VL measurements 3-12 months after estimated SC. Using multiple regression models, we examined whether initial VL and VLSP had changed by year of SC (≤1999,2000-01, 2002-03, 2004-05, 2006-07, 2008-9, ≥2010), adjusting for age at SC, SC interval, and time to first VL measurement. Time was modeled using restricted cubic splines to allow for non-linear temporal trends. Results: 1194 MSM seroconverting in median (IQR) 2006 (2003-2010) at median age 32.6 (26.8-40.1) years were included. They were predominantly white (91%), with SC interval 4.8 months (1.6-8.1), time to initial VL 2.9 months (1.1-4.5), and with VLSP based on 2 (2-3) VL measurements. Median (IQR) initial VL was 4.99 log10 copies/ml (4.34-5.62), increasing from 4.71 (4.05-5.21) ≤1999 to 5.05 (4.46-5.70) ≥2010. In univariate models, there was strong evidence of an increase in initial VL equivalent to 0.2 log copies/ml (95%CI 0.06-0.34) per SC year decade. After adjusting for other covariates, however, there was no evidence of such a linear trend (p=0.104), although initial VL ≥2012 was 0.31 log10 copies/ml (95% CI 0.09-0.54) higher compared to ≤1999. Median (IQR) VLSP was 4.64 log10 copies/ml (4.05-5.08) and increased from 4.51 (4.10-4.89) ≤1999 to 4.76 (4.15-5.16)≥2010. In adjusted models this was equivalent to an increase of 0.23 log10 copies/ml (0.05-0.42) over the 15 year period, but there was only borderline statistically significant evidence of a linear increasing trend over time (p=0.074). Conclusions: Our results suggest an increase in HIV virulence among UK MSM. This finding may translate to an increase in HIV incidence in the UK.