The relationship between sound duration and detection performance has long been thought to reflect temporal integration, and has been well studied in both humans and animal models. Reports of species differences are equivocal, with some metanalyses reporting no species differences, and others reporting substantial differences. This renders translational work in animals problematic. To re-evaluate this issue, tone detection performance was measured in rhesus macaques using a Go/No-Go reaction time detection task at various stimulus durations, and in the presence of broadband noise (BBN). Detection thresholds, reaction times (RT), and psychometric function slopes were calculated. All three measures were sensitive to tone duration, consistent with temporal integration. The rate of threshold change with duration was similar to human data, and was equally well fit by power law and exponential functions. The effect of tone duration on RT is the first reported in animals. BBN significantly affected how slopes changed with duration, and elevated thresholds, but did not affect how threshold or RT changed with duration. To begin investigating what processes may underlie these behavioral measures, the data were then compared to data generated using a probabilistic Poisson process model, previously compared to human data. The Poisson model recapitulated the effects of duration on threshold and psychometric slope. These behavioral data provide evidence that macaques are an exceptional model of human temporal integration, and the modeling results point the way forward for future neurophysiological studies.