HIV-1 p24 antigen (p24) measurement by signal amplification-boosted ELISA of heat-denatured plasma is being evaluated as an alternative to HIV-1 RNA quantitation in resource-poor settings. Some observations suggested that virion-associated p24 is suboptimally detected using Triton X-100-based virus dissociation buffer (kit buffer). A new reagent (SNCR buffer) containing both denaturing and non-denaturing detergents was therefore developed and evaluated. The SNCR buffer increased the measured p24 concentration about 1.5- to 3-fold in HIV-negative plasma reconstituted with purified HIV-1 particles, while not increasing the background. Among 127 samples of HIV-1-positive patients with moderate to high concentrations of HIV-1 RNA the increase was about threefold across the entire concentration range (P < 0.0001). Specificity before neutralization among prospectively tested clinical samples ruled HIV-negative was 828 of 845 (98.0%) for the SNCR buffer and 464 of 479 (96.9%) for kit buffer. Specificity after confirmatory neutralization of reactive samples or a follow-up test was 100% with either buffer. Surprisingly, the SNCR buffer revealed a p24 reactivity in 115 of 187 samples (61.5%) from adult patients exhibiting undetectable HIV-1 RNA below 5 copies/ml for a duration of 6-30 months under HAART (3.7% with kit buffer). The rate of p24 reactivity in these patients did not decrease with duration of HAART. In conclusion, the SNCR buffer improves the detection of particle-associated HIV-1 p24, thereby increasing the measured p24 concentration in samples with medium to high HIV-1 RNA. It also uncovers the presence of a p24 reactivity, whose identity remains to be determined, in a significant fraction of samples with undetectable HIV-1 RNA under long-term HAART.